Future of the PPC? (970, 980, 990, 9900)

[MacRumors]

The following information is from an unconfirmed and anonymous source. As such, authenticity is always uncertain, but due to the content of the piece, was felt to be of sufficient interest for publishing. Of interest, MacBidouille has posted (http://us.macbidouille.com/niouzcontenu.php?date=2003-07-05#134) similar details in their unconfirmed rumors of the PPC's timeline. This may represent corroboration -- or simply a common source. Take, as with all rumors, an appropriate amount of skepticism.

Apple and IBM have been working on parallel development of the Power5 and PPC 980. The PPC 980 is a single core version of the Power 5. While prototype forms of this chip exist, it is almost a year away from shipping in Macs.

Improvements in the PPC 980 include Hyperthreading, eLiza error correction, and more massive parallelism. IBM's implementation of hyperthreading provides a 30% gain over Intel's. eLiza technology will reduce the bottlenecks when the branch prediction unit fails. Altivec will split into 3 pipelines (vs 2 in the 970), 4 Integer and 4 Floating point units. 980 will have to be built on a 90nm processor due to heat dissipation requirements.

Steve's comment of 3GHz in 1 year will not be accomoplisedh by the G5 - which will top out at 2.6-2.8GHz. The PPC 980 will start at speeds of 2.6-3GHz and top out around 4.5-5GHz. The G5 will make it's way into PowerBook lines in Jan/Feb, Xserve's later this year, and iMacs in approximately one year.

Marklar's project size has decreased, but remains ongoing. There are four generations of the PowerPC including and beyond the 970 that are in development and planning. Besides the 980 chips (targetted at end of 2004), there are plans for 990 chips on a 65nm process in 2005/2006 @ 6GHz and scaling up to near 10GHz. Beyond this, the PPC 9900 starting on a 45nm process is targetted in 2007/2008 starting at 9-10GHz and reaching up to 20-25GHz by 2010-2011.

[arn]

This could be true... it could be made up. I don't know. But is an interesting read... so enjoy.

I guess it's something to keep in mind as more details surface over the next year.

arn

[k2k koos]

it's good to see they are not resting on their b*tts, but keepp on developing products for the future, to stay current, and perhaps even widening the performance gap...

[evil]

20-25 ghz
hot damn

[QCassidy352]

good lord, makes me not even want a g5 anymore! ;) I hope it's all true. That sounds great.

One question though- can someone explain "power 4" and "power 5" as compared to 970 and 980? The 970 is a power 4 derivative, while the 980 is a single core power 5? Is the G4 also a power 4 derivative? Is the 990 also a power 5 derivative? Thanks for the help.

[ZildjianKX]

Well, some interesting PB news. It would really suck for the towers to go to the PPC 980 chip and the PBs to still be G4s... I guess I'll have to hold out on the PBs for another 6 months... I think it will be worth it in the long run, maybe a new ipod by then anyways... not to mention preloaded with Panther.

[Ja Di ksw]

Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much? I guess that's what the iMac's and the like are for, though by then they will have 10 or 20 Ghz or whatever.

Remember this? :)

"640K ought to be enough for anybody"
- Satan, um, I mean, Bill Gates.

[Kermit]

I'm no fan of either Bill Gates nor Microsoft but that "640 kb is good enough for everyone"-quote is false. Bill Gates never made such a remark and it is time that we lay that myth to rest.

[macdop]

I am no expert on chips etc., but a 45nm sounds a little imposable, as I believe that would mean like 3-4 atom gates, which I believe is imposable even with UV fabrication technologies, and would also cause random electron jumps from gate to gate hence making transistors imposable, as I said, I am no expert, and I could very well be wrong, and I have just come back from a bar, and am very drunk, but I think simple physics makes this impossible

[GregA]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? <snip>

Remember this? :)
"640K ought to be enough for anybody"
- Satan, um, I mean, Bill Gates. Ah, 10 years from now, a friend walks in and says "nice garden out your window" - you say "no, actually that's my 3d Mac interface, with garden themes - pull out that carrot and watch what happens".

Or maybe they'll use the 25Ghz for a complex AI which only weeds out Spam and ads.

[Veldek]

I'm quite sure that I read some days ago, that IBM has a 970 working at 3.2GHz dissipating 82W, which is as much as the Intel chips at the same frequency. So it doesn't seem impossible for them to reach the 3Ghz barrier with a 970.

Anyway, as IBM is working on the Power5 (the 980 will be a derivative of it, the G4 has nothing to do with either of them) and on the Power5+, I think we may expect a lot in the near future.

[C14ru5]

What's the point of having eLiza error correction in a consumer processor? I thought error correction only was necessary in server processors like Power4 and Power5. I mean, how often do calculation errors occur in a processor chip?

[BaghdadBob]

20-25 Ghz in four-five years?

Wow.

I wonder how true that is.

To anyone who wonders what you would do with all that power, just remember that the more power your computer has, the smarter it can act. There are many things computers all do today that never would have been practical on sub 100 Mhz processors.

Trust me, if the power is there, someone will find a way to use it. Like Longhorn and the wavy window effect!

Actually, I'm waiting for the day when my desktop can fluidly animate like in Lawnmower man.

[LostPacket]

Originally posted by Macrumors
Beyond this, the PPC 9900 starting on a 45nm process is targetted in 2007/2008 running up to 20-25GHz.

Correct me if I'm wrong, but at 25 GHz an electron will only have 4E-11 seconds to complete a cycle and could travel about 1.1cm (assuming 90% speed of light).

How large will these chips be? Will the electon even be able to get across the CPU in time to complete its cycle?

I don't know anything about chip design, but these frequencies seem a little far fetched to me.

[MyLeftNut]

20-25Ghz?

Wow...you know what will happen though, your new PowerMac becomes self aware as of 12.35:48pm on Jan 02, 2009 and then we all vaporize in a blinding flash of light....

Sorry couldnt help myself...

[h'biki]

Yeah, I bet Steve will love to announce the QMac, the world's first personal quantum computer.

The comes the aiMac, the world's first artificially intelligent, fully aware personal computer :)

And then Steve downloads his consciousness into the computer and calls himself Steveomancer.

[midifarm]

Wahoo! You're all clear kid. Now blow this thing and let's get out of here!

It'll be a wonderful day when even the ./ers will wonder why they ever thought Intel would do the job. Go Big Blue and the Core! :D

[sososowhat]

All the physics aside (and way beyond my expertise), so long as Moore's Law continues, there's no news here.

Doubling every 18 months (yes, I know it has to do with transistor counts, not GhZ, but historically it works for either):

3 years = 4x
5 years = 10x
10 years = 100x
15 years = 1000x
20 years = 10000x
30 years = 1 Million X

So, by 2033 your Mac should be running around 1 Million times the speed of a G5.

And - more to the point - in 5 years, it should be running at 20 GhZ, which is what this "breaking news" says.

1986, my first PC:

6 MhZ 286
1 Meg RAM
30 Meg disk
1200 baud modem

15 years later (2001) - by Moore's Law 1000x - curiously it applies to everything, not just the chip:

6 GhZ chip (ok, we're off by 3 years)
1 Gig RAM (yup, right on)
30 Gig disk (right on schedule)
1.2 Mbaud(?) is about 1.5x my DSL

[LostPacket]

Originally posted by MyLeftNut
Wow...you know what will happen though, your new PowerMac becomes self aware as of 12.35:48pm on Jan 02, 2009 and then we all vaporize in a blinding flash of light....
I'd like to give the future race of machines the benefit of the doubt.

After all, any self respecting intelligence would spend its time searching for porn. Maybe a PowerMac G10 would search for a screencap of a dual-G5 with the side panel open ("Oh baby, look at those heat sinks").:)

I need sleep.

[mac15]

damn thats cool, some rather nice news and wow .65nm is so damn cool

[Shadow_Raptor]

First up QCassidy352 the IBM 970 is based on (or derived) from the IBM Power 4 server processor. The 980 is based on the Power 5 server processor. In both cases the Power 4/5 processors had dual cores while the 970/980 have only one core. They are a slightly less powerful processor with some changes made to suit their needs as a desktop prcoessor in Apple Macs.

The G4 was of course made by Motorola and so has nothing to do with the IBM Power 4 server processor.

And C14ru5 eLiza error correction is for when the branch prediction fails. The processor tries to guess what it will be asked to do - and begins to do that before it gets told to. If it gets it right it continues - if it gets it wrong it dumps the whole branch and starts from strach - very efficient if its right (which it is 80-90% of the time) but costly if it isn't.

As for everyone worrying about the 20-25Ghz 45nm process rest assured that this problem has been expected for some time now. It is true that the theoritical limit for current processors is up around 30Ghz. After that everything gets too small, too fast, too power hungry and too hot. Of course was to over come this will eventuate - think quad Power Macs - but I don't think we will see desktop computers going Quantum any time soon. I say this because Quantum computers are only fast at what they do - that is to say not very much.

[iGav]

Either way... I think IBM is very good for Apple!! :p :p :p

[Analog Kid]

Originally posted by QCassidy352
One question though- can someone explain "power 4" and "power 5" as compared to 970 and 980? The 970 is a power 4 derivative, while the 980 is a single core power 5? Is the G4 also a power 4 derivative? Is the 990 also a power 5 derivative? Thanks for the help.

IBM, Mot and Apple co-developed the PowerPC architecture so all of the chips share the same basic instruction set. This is why applications don't need to be recompiled when Apple changes manufacturers and why the Apple-Intel rumors were bunk from the start ("uh, now that you guys just finished porting your applications to OS X, we need you to port to an entirely different instruction set-- mind your byte ordering!!").

IBM and Mot cross-licensed the Alti-vec instruction set. Alti-vec is what made the G4 the G4 and has since been carried into the G5.

Power4/5 are IBM brand names, so the G4 (a Mot chip) isn't derived from either. Again, they all use the same instruction set (the G4, G3 etc use a 32bit extension to the original 64 bit PowerPC specification).

The 970 is a "stripped down" Power4. The Power4 is dual core, the 970 is single core. I think the Power4 also had a larger cache and an integrated memory controller which the 970 doesn't. The only addition I'm aware of is that the 970 added Alti-vec which the Power4 had no use for.

The Power4 is also a "beefier" more robust technology. Fatter oxides on the transistors and such for the never-fail requirements of a server. This actually slows the Power4 down a bit so slimming down the process allows the 970 to run at a faster clock rate than the Power4 could.

Ars Technica had a really detailed set of articles on the 970 if you're interested in that level of geekiness.

The 980 will make similar modifications to the Power5, except this time they are being designed in parallel.

All indications are that Apple and IBM will continue following this methodology and Apple will continue to use the IBM chips until Mot pulls their collective thumb out of their collective.. uh... ear... and realizes that cell phones aren't going to keep the company afloat.

This is the first mention I've seen of the other two chips, but I'd guess the 990 is a Power6 derivative.

If you're not interested in the fundamental architecture, I think the important thing to carry away from this is that IBM has an interest in developing world class processors for their own servers and workstations (which net much higher profit margins than just the chip would) so as long as the relationship between Apple and IBM remains warm and Apple doesn't start eating into IBM server sales we can expect leading edge processors in our Macs.

The deal is good for IBM because the added volume going to Apple keeps the IBM fab lines running more efficiently and helps defray the R&D costs. It's really a win-win.

Mot never had this incentive. Apple isn't going to build their Mac's around a cell phone processor (and Mot doesn't even use Mot processors for their phones, which should tell you something) and the biggest application for Mot PowerPCs is routers and networking equipment which is a very different problem with very different design goals. Beyond that it's mostly "embedded" applications like automotive.

IBM has an insane R&D budget and they span everything from practical manufacturing and OS development to basic physics research that won't show up in a product until, well, until the 9900 :)

[jbrown]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much? I guess that's what the iMac's and the like are for, though by then they will have 10 or 20 Ghz or whatever.
.

I remember pushing the envelope at 8 mb RAM:rolleyes: :rolleyes:

[skunk]

Does 25GHz mean I'll be able to type faster?

[yossele]

It'll type before you decide what to type! LOL.

[jbrown]

Originally posted by skunk
Does 25GHz mean I'll be able to type faster?

Although you'll be assisted by OS XXXIV ( codename TIDDLES ) inbuilt pre-emptive ' watcha-wanna-type' technology ( see system preferences ) that theoretically will type your report about a week before you even knew you had to type it - this advantage will be completely ofset by the fact that we will all be forced to use M$ Office 9,000,000.1.2a, thereby causing you to be late with your report--again

ah progress.......:D

[jbrown]

Originally posted by yossele
It'll type before you decide what to type! LOL.

Damm...beat me to it......I must learn to type faster!!:mad:

[Cochrane]

Well, about the 20 Ghz thing: I would have my doubts even if it was an official press release by IBM and Apple. Especially in the computer world the time from now to 2009 is very long and the physical aspects discussed above will require big changes. Anything can delay 20 Ghz or promote, or maybe, due to new or current technologies, we won't be measuring computer speeds in frequency anymore but maybe in how many calculation units you have. The further in future they predict the less reliable the rumors get, I think, just because so many things can happen even if it was true at the beginning.

[Analog Kid]

Ah, just a couple posts too late on the processor comparison... Maybe if I could express myself more succinctly...

Yeah, 45nm poses all kinds of physics headaches, but then so did 180nm and 130nm... Every year the industry talks about how Moores Law is about to hit a wall but then they have another session in the same conference explaining how to beat the limitations.

People never thought we'd be able to get this small with optical lithography and we have.

What is becoming the limiting factor is cost. I don't even want to venture a guess at what the mask costs for a 45nm chip will run. This is why we're seeing supposed competitors team up to build fabs and co-develop technologies-- they can't afford the investment risk to go it alone.

It's also why it's such a good idea to start using IBM as a supplier. Mot seems to be having trouble getting down to 90nm whereas IBM is likely to be one of the last companies standing in the race to smaller processes.

Even the great Intel is usually a step behind IBM in pure technology...

The G5 meant a lot more than just a faster PowerMac. It meant a roadmap for the future, which is oh-so-much more important.

Finally, don't think of 25GHz in terms of doing the same thing faster. Think of it in terms of shorter development cycles for new software. Assembly coding is quickly going the way of Latin. Look at how quickly amateur coders are putting out Cocoa apps. If you have more cycles to burn, you can focus less on optimizing and more on putting apps together quickly with more reusable code.

[Sabenth]

Just going to ask this whats the value of 25ghz as far as costs go how much would such a chip cost...

Taking into acount that yeap 25 ghz chips can and will be around and if you use a 1.4 ghz chip now and its slow well sure 25ghz will be slow one day too..

bring on the fluid content screens and anamations and voice over stuff oh and bring in some of that star treck stuff for recreating food ..

[Centris 650]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much? I guess that's what the iMac's and the like are for, though by then they will have 10 or 20 Ghz or whatever.

When I had my Centris 650(a 25 mhz machine) My work a few years later purchased a new 601 machine. I was stunned. I didn't think we'd ever need to upgrade since it was so fast.

Just saying that applications and the OS will take full advantage of the speed of the processor....eventually.

[Analog Kid]

Originally posted by Cochrane
Well, about the 20 Ghz thing: I would have my doubts even if it was an official press release by IBM and Apple. Especially in the computer world the time from now to 2009 is very long and the physical aspects discussed above will require big changes. Anything can delay 20 Ghz or promote, or maybe, due to new or current technologies, we won't be measuring computer speeds in frequency anymore but maybe in how many calculation units you have. The further in future they predict the less reliable the rumors get, I think, just because so many things can happen even if it was true at the beginning.

It probably came from a marketing roadmap rather than anything concrete. Some wonk sat down and started doubling every 18 months and decided that's the clock speed they'd need to stay competitive.

My question is when they're going to dump the clock all together and just go with asynchronous self-timed processing.

At some point that's got to be easier to design and it would require so much less power...

We'll probably see it out of TI first in one of their DSPs and then it'll filter through others. Eventually Intel will change the name, claim they invented it, and bring it to the mainstream. (EPIC, hyperthreading?!)

[Mr. EMan]

Infineon has cranked it up to 110 GHz already (well, not shipping, and not a micro).

Register article link (http://www.theregister.co.uk/content/3/31561.html)

[jbrown]

..from Macworld ( UK )

Product managers: G5 '4.3 billion times better than G4'

By Simon Jary

A group of Apple product managers met in London yesterday to discuss Apple's WWDC announcements: the Power Mac G5, Mac OS X 10.3 Panther, and iChat AV. These included: Tom Boger, Director of Power Mac Worldwide Product Marketing; Chris Bourdon, Senior Product Line Manager, Worldwide Product Marketing; Souad Laoussadi, Desktop Marketing Manager, EMEA; and Stephen Kelly, Pro Software Marketing Manager, EMEA.
Product managers on OS X 10.3.

A billion billion big numbers

Boger started his look at the new Power Mac G5 by running through the "giant leap" in processor technology made by IBM and Apple.

The PowerPC G5 is the PowerPC 970 chip, which is based on IBM's Power4 processor – which Boger called "the world's most sophisticated processor". The G5 is the size of an adult's fingernail and yet includes 300 metres of wiring; the circuitry is 800 times thinner than a human hair.

Boger was keen to point out that the G5's 64-bit architecture is not merely double that of the previous 32-bit chip builds. As 32-bit means 2 to the power of 32, and 64-bit means 2 to the power of 64, the new architecture is actually 4.3 billion times as expandable.

For example, although Apple states the maximum memory of the top-end Power Mac G5 as 8GB, it could - if the DIMMs existed - stretch to an almost unthinkable 18 exabytes (a billion billion bytes). 1 exabyte equals 1,024 petabytes. 1 petabyte is a quadrillion bytes. 18 exabytes is a hell of a lot of RAM!

Boger claimed that if memory requirements doubled every year, the G5's 64-bit architecture would still support our needs in 32 years.

In a similar fashion 32-bit computers can express numbers up to 4 billion. 64-bit can reach up to 18 billion billion. In the short term this power will be restricted to such fields as scientific research and cryptology, but will eventually filter down to more desktop-orientated operations. Yet the G5 can still happily run today's 32-bit applications, even those running in OS X's Classic environment. This is because when Apple, IBM and Motorola designed the original PowerPC back in the early 1990s, it was always planned to be a 64-bit chip - unlike Intel's Pentium. The Pentium's 64-bit successor, the Itanium, is "years away from reaching the desktop" claimed Boger.

Panther will remain 32-bit, however, as nearly all that power is still only needed at the very highest levels of computation. Developers can still write calls that take advantage of the 64-bit architecture. And Apple has rewritten the way that 10.2 Jaguar addresses memory. Mac OS X 10.2.7 - which ships in the G5 - has no memory limit, allowing up to 4GB of RAM per processor using the DIMMs that are available today.

Cool case
Boger praised the Power Mac's efficiently engineered aluminium enclosure – the number-one goal of which was acoustics. It has a fine-tuned thermal management, including an Apple-invented fan-control system that means a 5,000rpm fan can spin down to a much quieter 500rpm depending on the operation being carried out by the user. The Power Mac director admitted that the cooling system is designed for each model, and therefore Apple does not recommend later processor upgrades in the new systems.

The transparent air deflector has scoops for channeling air through the enclosure. If it is removed, the Power Mac automatically sends itself to sleep because the air-cooling system will be compromised. As soon as it is replaced, it wakes itself up again.

No tools are required to swap out either optical or hard drives, and Apple even includes spare "buttons" for users to add to third-party external hard drives so that they fit like that already installed in the top drive bay.

[jbrown]

NOTE THIS BIT ON MY LAST POST---

The Power Mac director admitted that the cooling system is designed for each model, and therefore Apple does not recommend later processor upgrades in the new systems.

---Surely a way 'round this me'hopes

[Nermal]

I've only had a quick browse through this thread, but I haven't seen anyone say "but they only just released the G5, they can't release the G6 in just one year, it has to be further away than that". If anyone is thinking about this, remember that a 980 can still be called a G5. Just like how we went through several Motorola 74xx chips that were all called G4.

I guess I'm being a bit pre-emptive, like OS XXXIV's auto-typing function mentioned above, I'm answering the question before anyone asks it :D

[pwm519]

[B] "All the physics aside (and way beyond my expertise), so long as Moore's Law continues, there's no news here.

Doubling every 18 months (yes, I know it has to do with transistor counts, not GhZ, but historically it works for either):

3 years = 4x
5 years = 10x
10 years = 100x
15 years = 1000x
20 years = 10000x
30 years = 1 Million X

So, by 2033 your Mac should be running around 1 Million times the speed of a G5.

And - more to the point - in 5 years, it should be running at 20 GhZ, which is what this "breaking news" says.

1986, my first PC:

6 MhZ 286
1 Meg RAM
30 Meg disk
1200 baud modem

15 years later (2001) - by Moore's Law 1000x - curiously it applies to everything, not just the chip:

6 GhZ chip (ok, we're off by 3 years)
1 Gig RAM (yup, right on)
30 Gig disk (right on schedule)
1.2 Mbaud(?) is about 1.5x my DSL" [B/]



A quick look and I think the 6Mhz 286 had 29,000 transistors. Todays P4s have 42 million transistors. In 15 years we should have had 10 (18mo) periods or we should be at 30 million transistors? Are we actually a head of Moore's law?

[Veldek]

Originally posted by skunk
Does 25GHz mean I'll be able to type faster?

No, it means you HAVE TO type faster to take use of the whole power this thing offers. ;)

[CmdrLaForge]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much? I guess that's what the iMac's and the like are for, though by then they will have 10 or 20 Ghz or whatever.

Remember this? :)

"640K ought to be enough for anybody"
- Satan, um, I mean, Bill Gates.

Hi,

I don't know as well. I even don't know for what I could need a G5. At the moment I have a iBook with 900MHz. It does all I want at a speed that is quite enough for me.

But on the other side - I want to order a G5 (middle system) - maybe just to have one.

And if they offer a 25Ghz system - I will buy it.

Cheers

[xelterran]

im guessin this thing is going to have about 100TB of ram..?

[MRLurker]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much? I guess that's what the iMac's and the like are for, though by then they will have 10 or 20 Ghz or whatever.

Remember this? :)

"640K ought to be enough for anybody"
- Satan, um, I mean, Bill Gates.

Just think about all of the stuff that is computationally expensive to do now. Real time language translation for the web and IM, full 1024 bit filesystem encryption, Java applets loading in less than 5 min, yahoo maps not just giving you the approximate travel time but doing a full particle simulation on the air between your house and the Mall to get the travel time (well maybe not).

[MRLurker]

Originally posted by MyLeftNut
20-25Ghz?

Wow...you know what will happen though, your new PowerMac becomes self aware as of 12.35:48pm on Jan 02, 2009 and then we all vaporize in a blinding flash of light....

Sorry couldnt help myself...

Unfortunately it was an accident, when the PowerMac became self-aware it jostled the PC next to it and knocked the heatsink off of the Athlon (Mr. Fusion edition) inside. Losing the heatsink on the Athlon is what caused the flash. If you don't know where this joke comes from just follow the link:

http://www.tomshardware.com/cpu/20010917/heatvideo-04.html

[Mr. Anderson]

Beyond this, the PPC 9900 starting on a 45nm process is targetted in 2007/2008 running up to 20-25GHz.

Nice - that might actually get me realtime radiosity rendering for Lightwave - something I've been continually waiting for. And I always thought it would be about 5 years away, this seems to be about right - if its true....:D

I wonder if these HAL 9900s will be multiprocessor.....:D

You know you shift each letter 1 place in the alphabet and you change HAL to IBM......

D

[mk_in_mke]

http://www.digitalproducer.com/cgi-bin/getframeletter.cgi?/2003/07_jul/features/07_01/cw_hp_ita2int.htm


check this article... nothing surprising but very interesting


Michel

[soggywulf]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for?

Real time computational fluid dynamics in my flight sim. Real time semi-rigid body calculations, same purpose. True translucent cloud rendering and other better graphics, although for this we need ATI and 5dfx to keep up with the roadmap.

[soggywulf]

Originally posted by MRLurker
Just think about all of the stuff that is computationally expensive to do now. Real time language translation for the web and IM, full 1024 bit filesystem encryption, Java applets loading in less than 5 min...

Nice. :D

[daRAT]

Originally posted by Kermit
I'm no fan of either Bill Gates nor Microsoft but that "640 kb is good enough for everyone"-quote is false. Bill Gates never made such a remark and it is time that we lay that myth to rest.

Ahh yes he did, do a web search for that phrase. All the famous quote sites list him as the author. He made it in 1981.

And at that time, it WAS enough :P

[adamfilip]

well lots of users now are running a gig of ram now in 2007
10 gigs? should be about average
with

4gig memory modules runniing in the $80 range.. QDR Quad data rate

also.. one thing.. if the chips runnig at 20ghz.. then the bus is running at 10ghz!.

i sure hope they make some drastic improvements on the price/storage ratio of solid state storage

other wize the bottleneck we have now will be crazy in a few years

[daRAT]

Originally posted by skunk
Does 25GHz mean I'll be able to type faster?

No your brain will be wired for Bluetooth, therefore your thoughts will be processed the the computer ...

[hvfsl]

With all these comments about typing on Macs, it seems most people only use their Macs for Office. I use my PC for Office and my Mac for creative stuff like PhotoShop. I thought that was the whole point of getting a Mac, doing creative stuff on it.

Anyway I can't wait for the new 20Ghz chips, I will be able to encode DVDs in realtime (although it will be C3D by then) and play games with better graphics than The Matrix Reloaded or Star Wars. Another tec I am really looking forward to is the special wallpaper that can act as a computer screen, so I can change the colour depending on my mood. It is based on the tec for the new flat screens (I think its called OLED).

[jouster]

As Soggywulf notes, there are many applications for which the 20 - 25 Ghz range would be necessary - real time rendering in games, for example.

Software companies will ALWAYS come up with apps that need more power to run. Of course, whether we need them to make our computing experience better or not is open to debate, but that doesn't matter - they WILL be developed.

Also, don't forget that the economic models on which hardware manufacturers base their developments depend on continuous speed and memory gains. People _must_ upgrade continually for the hardware companies to survive.

Finally, don't rule out the (probable) development of some sort of previously inconceivable application for so much computing power. These speeds will happen. Whether they happen in this article's timeframe is open to debate, but they _will_ happen.

[cooter]

i dont know how true this time line is, but i have read independent stories on cnn etc about the future of cpu's and they predicted that PCs (probably meaning Intel etc) would be in 10 ghz ranged around 2006-2007. and that by 2015 or so, we will have weather forecasts that can forecasts weather based on models accurate to 2sq feet and up to 30 days in advance. the weather guy in story said that would be a "holy grail" of forecasting.

things like simple AI i imagine will be more possible, a pc which knows it users mood etc thru face recognition or voice analysis. i can see games like warcraft having all the grunts have thier own little AI's like WETA's Massive.

while this is all good, i think this is bad for computer industry on the back-end, because honestly, most people dont even need a 1 ghz computer. a 5 ghz pc/mac could very well be the last computer you ever need to buy.

[rickmoen]

I wonder if Motorola will be up to 2 Ghz by then?

[BigJayhawk]

Originally posted by cooter
a 5 ghz pc/mac could very well be the last computer you ever need to buy.

OK, Bill Gates. (Too bad you don't have the money, too. Right?)

[Malic]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for?

Simple: Games. Games always eat the whole machine up. I'm sure Carmack will come up with something that test a 20Ghz processor to it's limits. :D

[jouster]

Originally posted by rickmoen
I wonder if Motorola will be up to 2 Ghz by then?

Yeah, we'll have a G8 at 20 Ghz with a 10 Ghz bus and multi Terabyte drives........and we'll STILL need a 56k modem for faxes!! :)

[soggywulf]

Originally posted by cooter
while this is all good, i think this is bad for computer industry on the back-end, because honestly, most people dont even need a 1 ghz computer. a 5 ghz pc/mac could very well be the last computer you ever need to buy.

This same statement could have been made at any point in the history of the PC. Substituting smaller numbers, of course.

[Frobozz]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors.

Answer: anything that requires massive amounts of math. Some possibilites:

-- Calculation of complex mathematical systems, including accurate physics.
-- Complex artificial intelligence. You'd probably need at least a hundred 25 Ghz 9900's to equal the computational power of a human brain: it's not just about computing an answer you're given, it's figuring out the question that's tough.
-- 3D rendering could be done in realtime with caustics, raytracing, etc.
-- Video composition.

It's all about being able to scale up the number of accurate mathematical processes.

[Macmaniac]

I want one already! Whats nice about IBM is that they are constantly updating their chips in speeds that our PC brethern enjoy. So we will be able to catch up and surpass the WinTel world!
I can't wait to play Warcraft 4 and 5 one these machines:)

[rjwill246]

Originally posted by soggywulf
This same statement could have been made at any point in the history of the PC. Substituting smaller numbers, of course.

And that is true of everything. Now, if only the software didn't suffer the inevitable bloat as well. Imagine the speed boost we would really get if efficiency came back into software writing. Machine.... Assembler...

Holographic images that don't look like see-through glass, communicating using AI and full speech recogntion with no delays in responses - fabulous if we could get back to efficient programming.

[jouster]

>>> And that is true of everything. Now, if only the software didn't suffer the inevitable bloat as well. Imagine the speed boost we would really get if efficiency came back into software writing. Machine.... Assembler...

Holographic images that don't look like see-through glass, communicating using AI and full speech recogntion with no delays in responses - fabulous if we could get back to efficient programming.


Hmmm....agreed, but don't hold your breath. Coding is only going to reach higher levels of abstraction to make it easier....think Xcode. The days of hand coding and assembly language are in the past and receding fast (hey, I'm a poet!). Bloat is here to stay unfortunatley, but if some of the awesome apps being predicted in this thread come to pass, so what? Also, if the processor can handle it, again so what?

[Pete_Hoover]

I am getting a new computer anywhere from august 2004-January 2005. Maybe I can get a rev. A 980 right when they come out. Along with those new rumored aluminum displays, that would be a pretty sweet setup.

[jouster]

Pete, I'm looking at the same time frame....see you in the store!

Though part of me is tempted to see how long my (modest) computing needs can be fulfilled by my current setup (PB G3 320 20), the other part is rapidly becoming a techno ho!

Edit: Holy status upgrade! I just became a member......nearly 18 months after I registered!

[MacFan25]

I wonder what speeds Intel and AMD will be running at if IBM hits 25Ghz. By that time, we will probably be loading our computers up with 200GB of RAM, too. ;)

[rog]

Apple is clearly not interested in increasing market share. They need to get the G5 in consumer machines imemdiately. They cost less apparently than the G4s, so why not? High end P4s and Athlons can be found in consumer level PCs for dirt cheap. 2 years ago the iMac had a 700 MHz G3. Not it's up to a whopping 1 GHZ, albeit with a crippled no L3 cache G4. A year to get a G5 is too long. What Apple needs to do is get low GHz G5s in the iMac, and put duals in all the towers. The entry price for a G5 is $1999 (although they can be had for $1799 with combo drive, the media isn't reporting this. Apple has serious marketing problems. A $1299 iMac should be able to keep up with a $799 PC. Even with a 1.4GHz G5, I am not sure that would be the case. With a processor half as fast, like they are now, the iMac, iBooks, and even PowerBooks are overprice, underpowered jokes. At least they look nice.

[jouster]

Well, this is because Apple sees them as consumer machines. The thinking is that if they run WP, browsers and the iApps efficiently then they are filling their market niche.

Also, there may currently be yield issues with the G5.

[Rincewind42]

Originally posted by Analog Kid
Finally, don't think of 25GHz in terms of doing the same thing faster. Think of it in terms of shorter development cycles for new software. Assembly coding is quickly going the way of Latin. Look at how quickly amateur coders are putting out Cocoa apps. If you have more cycles to burn, you can focus less on optimizing and more on putting apps together quickly with more reusable code.

Coding in assembly is going the way of Latin. Being able to read assembly however is still the hallmark of a good programmer on his platform. There will always be a need to optimize your code because people will always expect things to be faster. So, you can calculate pi to 10 million digits in 2 seconds? Why couldn't you do it in 1 second while decrypting a 10 GB file? Sure, it's great to be able to put an app together without writing more than a handful of code - but if you can do it so can thousands of others and your app really isn't all that distinctive in that sense then is it? A distinctive app requires a programmer to take full advantage of all of the tools available to him or her - including optimizing.

[rog]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much?

Apparently 25 GHz is needed to be able to resize a window in OSX as fast as my 25 MHz first mac could do it in OS 7.1 (an LC3). The GUI in OSX has essentially 10% of the performance of OS9. Such as with scrolling or the let 1000 windows bloom benchmark. So, until there is a 12.5 GHz mac running OSX, a 1.25GHz mac running 9 will still feel more snappy. Very sad.

[Freg3000]

Originally posted by jouster
Yeah, we'll have a G8 at 20 Ghz with a 10 Ghz bus and multi Terabyte drives........and we'll STILL need a 56k modem for faxes!! :)

And PCs will STILL have Serial and Parallel ports. What is worse? :)

[Jeff Harrell]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for?By 2005, you're going to be shooting your home movies in uncompressed high definition video and editing them to HD-DVD in real time.

Current systems can barely keep up with DV video. We have a long way to go yet before we can talk about how nobody needs more oomph.

[Jeff Harrell]

Originally posted by daRAT
Ahh yes he did, do a web search for that phrase. All the famous quote sites list him as the author.Oh, well then, it must be true.

[Jeff Harrell]

Originally posted by Frobozz
-- Complex artificial intelligence. You'd probably need at least a hundred 25 Ghz 9900's to equal the computational power of a human brain.AI is not a computationally bound problem. It's not like we have AI systems out there right now that can do the job but are too slow. AI is a theoretically bound problem: we don't even know if it's possible yet, much less if it's possible with existing technology, much less whether existing technology is fast enough.

[Rincewind42]

Originally posted by rog
Apparently 25 GHz is needed to be able to resize a window in OSX as fast as my 25 MHz first mac could do it in OS 7.1 (an LC3). The GUI in OSX has essentially 10% of the performance of OS9. Such as with scrolling or the let 1000 windows bloom benchmark. So, until there is a 12.5 GHz mac running OSX, a 1.25GHz mac running 9 will still feel more snappy. Very sad.

When System 7 came out people were complaining about how much slower it was than System 6. Never mind that it did things that System 6 could/would never do. So Mac OS X is slower than Mac OS 9 for a few things - so what. It can do things that Mac OS 9 will never do.

[Rincewind42]

Originally posted by rog
Apple is clearly not interested in increasing market share. They need to get the G5 in consumer machines imemdiately. They cost less apparently than the G4s, so why not? ... A year to get a G5 is too long

Apple is intensly interested in increasing market share. But that doesn't necessarily equate into G5 iMacs in less than a year. With the G5 in the PowerMacs Apple can now put out 1.4 Ghz iMacs (although I doubt that will be soon). Yes it would be nicer if they were G5s now - but looking at the engineering that went into the PowerMac G5 I can probably tell you right now that it will take a year to get them into the iMac form factor without them sounding like a Windtunnel G4. I think it has very little to do with cost (which no one actually knows at this point) and just simple engineering issues.

Originally posted by jouster
Also, there may currently be yield issues with the G5.

I don't know for certain, but I doubt there are yield issues. The PowerMac G5s at WWDC were preproduction models - not a one off the production line. It seems that Apple pretty much said 'Is the design done? Ok, we'll announce it'. And that is why they won't ship till August (or early September). They have to have enough machines to fill the market place too.

[jouster]

Originally posted by Freg3000
And PCs will STILL have Serial and Parallel ports. What is worse? :)

Hmmm. I dunno. I am embarrassed to admit that I don't know what serial or parallel ports are!!!!

[Flowbee]

Originally posted by Macrumors

Besides the 980 chips (targetted at end of 2004), there are plans for 990 chips on a 65nm process in 2005/2006 @ 6GHz and scaling up to near 10GHz. Beyond this, the PPC 9900 starting on a 45nm process is targetted in 2007/2008 running up to 20-25GHz.

25Ghz is all well and good, but the important question is... what will the *case* look like? :p

[P-Worm]

So the FSB is supposed to scale with processor right? So does that mean we will be seeing a 12.5 Ghz FSB in the near future?

All I can say about that is :cool:

P-Worm

[gwuMACaddict]

mmm... those predicted speeds sure do sound tasty. i'm a little dissapointed about no 3ghz g5, but it sounds like the 980 will more than make up for it.

now, where to find the money...

:D

[mania]

regarding 25 GHz I remember back in the OS 8 days I think it was Connectix had products called RAM Doubler and Speed Doubler. Someone posted a fake product called Brain Doubler which said something like; "99% of the time your CPU is waiting for you to do something - now you can make use of that wasted time with Brain Doubler..."

[Phinius]

It could be that IBM will advance the core frequency of the 9XX according to when faster main memory is available. If you look at the new Power Macs listed on Apple's store you'll notice that the G5s bus runs at half the speed of the processor core. Well if Apple/IBM want to take full advantage of faster bus speeds they need to wait until faster memory is available before advancing the frequency of the processor. Currently, 400 MHz DDR memory is the fastest memory manufactured in quantity for personal computers. Later this year 533 MHz DDR-II memory should appear and the 970 would likely use this to advance to 2100-2500 GHz. In order to make full use of the bus when the 970 gets to 2.6 GHz in the middle of 2004, Apple would need to use 666 MHz DDR-II memory and perhaps not coincidently 666 DDR-II memory is not expected to be mass produced until the second half of 2004.

It seems to be much more difficult to advance the memory bus speeds on the Pentium 4 or Opteron processors than it is for the 970 chip. The Pentium 4 processor has only had three different DDR bus speeds since it was introduced. Whereas Apple already lists different bus speeds for all three Power Mac G5 computers. It's likely that Intel will not move beyond 533 MHz DDR-II memory in 2004, so Apple should retain the fastest memory bus speed for a desktop computer at least through 2004.

[XnavxeMiyyep]

Originally posted by jouster
Hmmm. I dunno. I am embarrassed to admit that I don't know what serial or parallel ports are!!!!
I dunno either. They're probably something already outdated. I think they still make PC's that use pre-USB ports to hook up the keyboard.

[macrumors12345]

Originally posted by CmdrLaForge
Hi,

But on the other side - I want to order a G5 (middle system) - maybe just to have one.

Cheers

Wow, Geordi, you are the first person I have seen who is planning to buy the 1.8 Ghz model. I have seen a number of posts saying "I plan to get the Dual 2 Ghz G5," but this is the first one I've seen saying "I want to get the Single 1.8 Ghz model." Good to know that someone is buying them. Still haven't seen anyone going for the 1.6 Ghz.

[actripxl]

The down side is we have to wait another 6 months till the PB get an update! I wanted to get a PB but I refuse to pay close to $3000(including tax) for a laptop that is over 8 months old and even if it is updated will still have a obsolete G4. If Apple wants me to buy a G4 laptop I will but they gotta update it and knock off at least $200. This may sound selfish, but I don't think OSX is worth THAT much for me to pay such a premium for old hardware.

[Yosh]

There's an inconsistency in this article.

1) Steve says we will reach 3 GHz in the next 12 mos (say by Aug-Sept 2004)
2) Article says 3GHz will NOT be accomplished with the G5 which will top out at 2.6-2.8GHz.
3) Article says the PPC980 will start at 2.6-3 GHz and top out at 4.5-5 GHz.
4) Article says the 980 chips are targeted at the end of 2004 (which I assume to mean Nov-Dec 2004).

Thus, the article contradicts what Steve had said at the WWDC Keynote (3GHz by Aug-Sept 2004). Either we will have a 3 GHz G5 (970 chip) or the 3 GHz 980 will be available earlier than the "end of 2004". In either case, throw in the nonsense about 20-25 GHz by 2007-2008 and the article doesn't seem to be all that realistic.

[cr2sh]

I was buying into the article until the "g5 in powerbook jan/feb" line, I simply don't believe it... itll be a year before we see those.

[Haberdasher]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much? I guess that's what the iMac's and the like are for, though by then they will have 10 or 20 Ghz or whatever.

Remember this? :)

"640K ought to be enough for anybody"
- Satan, um, I mean, Bill Gates.

You just contradicted yourself. Just how 640k used to be pushing the edge of sanity, eventually it became far, far too little. The same goes for 20-25 Ghz. You can bet, dollars to doughnuts, that by the time we are there with that much processing power, the applications we run will have kept up with speed requirements every step of the way.

And Virtual PC will still be slow as hell. :D

[macrumors12345]

Originally posted by C14ru5
What's the point of having eLiza error correction in a consumer processor? I thought error correction only was necessary in server processors like Power4 and Power5. I mean, how often do calculation errors occur in a processor chip?

Yeah, I'm very suspicious too. If you do a Google search for "eLiza AND IBM", you will find that eLiza is an IBM project that attempts to reduce the amount of labor needed to maintain servers by implementing "intelligent systems that self manage and regulate themselves." It's mostly implemented at the software, not the hardware level, let alone the microprocessor level. Basically, it's very hard to see ANY relationship between IBM's current eLiza products and some great technology to "reduce the bottlenecks when the branch prediction unit fails." This smells VERY strongly of being made up by someone who just concocted a story using different IBM technology buzzwords he found on the web, without having much idea of what they mean. I would give 10:1 odds at least that this story is false.

[Phinius]

IBM has announced that a POWER5 based computer is 4X faster than a Power4 computer. Intel has mentioned they are working on a Itanium processor that will be 10X faster than the currentedly produced Itanium processors and it will be manufactured in the middle of the decade. In order for IBM to keep up with that pace of improvement the POWER6 (expected in 2006) would need a 2.5X speed increase over the POWER5.

It would seem from the above that Apple has very good prospects for increasingly fast processors in the next few years. The company also stands a good chance of using much faster processors than what Intel produces for the desktop. Afterall, Intel intends to keep the performance of the Itanium processors at a higher level than the Pentium chips and IBM will have to match it with the POWER processors (9XX chips derived from POWER core).

[Mr. Dibbs]

Originally posted by macdop
I am no expert on chips etc., but a 45nm sounds a little imposable, as I believe that would mean like 3-4 atom gates, which I believe is imposable even with UV fabrication technologies, and would also cause random electron jumps from gate to gate hence making transistors imposable, as I said, I am no expert, and I could very well be wrong, and I have just come back from a bar, and am very drunk, but I think simple physics makes this impossible

okay, it's usually not my policy to reply without reading the thread, but I'm chiming in on this one. Intel has long had a 30 nanometer process in the works with a 2-3 atom gate. UV lasers aren't the only laser to use a invisible wavelength. think x-ray lasers.

[macrumors12345]

Originally posted by mk_in_mke
http://www.digitalproducer.com/cgi-bin/getframeletter.cgi?/2003/07_jul/features/07_01/cw_hp_ita2int.htm


check this article... nothing surprising but very interesting


Michel

I have no doubt that HP's new Dual 1.3 Ghz Itanic systems could give the Dual 2 Ghz G5 good competition computationally speaking (at least in FP intensive stuff), but there are not even close to being in the same market. An equivalently configured Itanic system will easily set you back $8,000, as compared to $3,000 for the Dual G5. Of course, it would be hard to actually benchmark the two systems against each other, since there is very little graphics or workstation type software available for Itanic.

Sounds like Charlie White doesn't think that Dual Xeons will be able to beat the Dual G5, so he is desperately seeking out other non-Apple machines that can match it. Too bad the ones he is finding can't run most of the software he's interested in benchmarking, and come at an incredibly high price.

[Abstract]

Originally posted by Phinius
....Well if Apple/IBM want to take full advantage of faster bus speeds they need to wait until faster memory is available before advancing the frequency of the processor. Currently, 400 MHz DDR memory is the fastest memory manufactured in quantity for personal computers. Later this year 533 MHz DDR-II memory should appear and the 970 would likely use this to advance to 2100-2500 GHz. In order to make full use of the bus when the 970 gets to 2.6 GHz in the middle of 2004....


Sure, or they could just release the units with 400MHz DDR RAM. Who cares if it works as long as it looks good on paper. :cool:

PS: Your computer will only be as fast as your slowest components. Thank goodness Dell stopped including floppy disk drives in their computers recently. ;)

[cooter]

i said "need" and "possibly" to buy, not want. i still think MOST people wont need a 5 ghz machine or feel a pressing need for anything faster . at very least hardware today is going further and longer between upgrades than hardware of 10 years ago. i dont see why that trend wont continue. im a graphic design and do professional print, video and 3d. i certainly want a 10ghz monster. but in all honesty, if i was limited to doing just print and typical home use, i can do everything quite happily at 1ghz speeds or lower. Same can be said for all my graphics pro friends, we all have been using machines up to 4 years old now for most things and none of us feel a pressing need for speed. im only one i know is who is thinking of even getting g5 amongst the 10 or so people i know in this biz. and thats because i do 3d.


as for the "same could be said 10 years ago" stuff i was offered in reply, well that was true then, but in honesty, what has been the killer app of last 10 years? email and web browsing. communication apps. this is still true even though the cpu speeds have gone from 60mhz to 3 ghz in that time. instant messaging doesn't need a 5ghz machine.

video games which had driven the cpu upgrade cycle in pcs' now drive the gpu cycle and console market more than cpu needs and having been accelerating at a faster rate. Microsoft even acknowledged this trend when it jumped into the Console market with the Xbox. no need for 5ghz pc or Windows if consoles are the future of home entertainment and perhaps more.
a new console at $300 every 4 or 5 years makes more sense than $1500 pc every 3.

the one thing i think WILL drive a need new for faster machines beyond 5 ghz will be real time voice translation. That will be the next killer app and will be water shed mark in human history.

[eric67]

Originally posted by Veldek
I'm quite sure that I read some days ago, that IBM has a 970 working at 3.2GHz dissipating 82W, which is as much as the Intel chips at the same frequency. So it doesn't seem impossible for them to reach the 3Ghz barrier with a 970.

Anyway, as IBM is working on the Power5 (the 980 will be a derivative of it, the G4 has nothing to do with either of them) and on the Power5+, I think we may expect a lot in the near future.

correct it was theinquirer.net or theregister.co.uk
actually the inquirer.net has a really nice article on the G5 and the 64-bit processors. a good one for Apple

[eric67]

Originally posted by macrumors12345
Wow, Geordi, you are the first person I have seen who is planning to buy the 1.8 Ghz model. I have seen a number of posts saying "I plan to get the Dual 2 Ghz G5," but this is the first one I've seen saying "I want to get the Single 1.8 Ghz model." Good to know that someone is buying them. Still haven't seen anyone going for the 1.6 Ghz.

this is so correct. In France, apparently the G5 dual is a hit, several hundreds of units have been ordered within the week following the announcement. but no news concerning the 1.6 and 1.8 GHz model. Actually macbidouille.com made a small article "buyer's guide" for the G5, and of course the low-end G5 is not a good deal (no PCI-X, RAM, FSB,.....)
now the question is why Apple did not propose a dual 1.6 or1.8 model, that would have been so good actually.... since OSX is taking finally full usage of dual processor why not making G5 by default dual processor??

[unreg]

IBM has used tunnelling electron microscope to place individual atoms to spell IBM in the 1980's. Intel announced 1 atom gates last year. Quantam computing advances are accellerating. 20-25 GHz machines are most likely in the next 10 years.
M$.02

[jbrown]

we are all of us decieved.......for there was another processor....one processor to rule them all, and in the darkness bind them......

but it needed too many fans!!:D :D

the treadmill of processor performance / application requirements will keep on running, and us paying -- it's good for business -- but not us

viva la revelution
viva carracho
viva libetad:D :D

[cuneglasus]

This artcle is rubbish.It cant even get simple facts about the 970 straight.The 970 has 4 altivec pipelines not 2.(IT has a simple integer,complex integer,floating point and permute unit just like the G4).Why do people want to publish such lies.

[phillymjs]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for?

If we get there first, to finally SHUT UP those Wintel "GHz-is-everything" fanboys and other Apple haters! :-)

~Philly

[Raiden]

I agree with whoever said that realtime voice translation will be the next killer app to hit the world by storm.

[Bonte]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much?

With this speed we can finally call a Japanese friend and talk to him in plain English. The software will translate in real-time and speak out in Japanese.
This everyday thing in the future will need some serious horsepower.

I'll bet there will be more. :cool:

[solafide]

It's demise was prematurely announced.

[solafide]

I remember when my brother bought his first computer and was told that no one would ever need more than 640k of memory.

We can't measure the future need for power by what is currently available, but rather by what we can imagine.

The example of real-time translation is a good one.

Does anyone remember Apple's vision for computing in their short 1987 video "The Knowledge Navigator"? It included an intelligent agent that created a more natural interface. I can't wait to get rid of the keyboard!

[Awimoway]

Good to see Marklar's still alive. It plays an important role in the checks and balances system going on.

As for the naysayers about 45 nm and 25GHz, I'm glad none of you work for Apple/IBM.

[unreg]

Don't forget carbon nanotube interconnects. No, it isn't in production yet, but shows great promise to bypass crosstalk at 45nm. (EEtimes.com?)
The computing power to do real time translation, high level AI, almost realtime universal db searches and direct man/machine interconect. (Kevinwarick.com)

[Frobozz]

Originally posted by Jeff Harrell
AI is not a computationally bound problem. It's not like we have AI systems out there right now that can do the job but are too slow. AI is a theoretically bound problem: we don't even know if it's possible yet, much less if it's possible with existing technology, much less whether existing technology is fast enough.

Well, it's both. But your point is taken.

[Nebrie]

Originally posted by Yosh
There's an inconsistency in this article.

1) Steve says we will reach 3 GHz in the next 12 mos (say by Aug-Sept 2004)
2) Article says 3GHz will NOT be accomplished with the G5 which will top out at 2.6-2.8GHz.
3) Article says the PPC980 will start at 2.6-3 GHz and top out at 4.5-5 GHz.
4) Article says the 980 chips are targeted at the end of 2004 (which I assume to mean Nov-Dec 2004).

Thus, the article contradicts what Steve had said at the WWDC Keynote (3GHz by Aug-Sept 2004). Either we will have a 3 GHz G5 (970 chip) or the 3 GHz 980 will be available earlier than the "end of 2004". In either case, throw in the nonsense about 20-25 GHz by 2007-2008 and the article doesn't seem to be all that realistic.

Read it again, slowly and carefully, it makes perfect sense.

[Phinius]

How high of a frequency a processor can achieve on a given process size has a lot to do with the amount of pipeline stages the core has. The 970 has 16 stages for integer-type instructions, 21 stages for floating-point instructions and up to 25 stages for Altivec (vector) instructions. Compare that to the Pentium 4 (20 pipeline stages) which will peak at about 3.2 GHz on the same process size as the 970 and it looks like it could very well be possible that the 970 will hit 2.6-2.8 GHz. If the person giving out this information to MacRumors is incorrect and just guessing, he or she is probably not off by much.

Look at it another way, the Power4 has 12 integer pipeline stages and the 970 has 16. The fastest POWER4 chip now runs at 1.7 GHz. Since the 970 has 1/3 more integer pipeline stages than the POWER4, then the 970 might be able to achieve 2.2+ GHz in the next few months. If this is in the realm of possiblity then I'd expect Apple to announce faster G5 speeds before the end of this year.

[asim]

don't stop at just 25GHz, because we'll need even more than this. I'm not even talking about translation... even some high quality voice recognition could eat up a large portion of this processing power.

I'm not just talking about a new version of viavoice, but real-time recognition that can do full speed transcription [edit: even with accents, which hampers the funcationality of viavoice for my parents], intelligent automated punctuation, and do so with the same hit to total cpu-load as a keyboard has on my imac g4/800.

add to that translation, another area where there can be orders of magnitude improvements. then, you have speech generation, from text, using preset voiceprints to automatically generate an appropriate rate, rhythm, and tonal inflection.

put together the speech recognition, translation, and voice speech generation and you have something which could potentially max-out a dual 25GHz.


what if you could hum a tune and it would automatically go through your itunes library and find that song, and if not find it from the itunes music store?

how about visual recognition? if you click on a picture of someone in a movie and it searches your iVideo (movie equivalent of itunes which stores your entire DVD collection) library to see that this star actor was the waiter extra in some movie 10 years before, which is where you recognize them from.

and some quality virtual-reality porn will require a quad GX (the "G ten") 42GHz with a 3d hologram projector... hopefully it will be ready by the time i find someone, get married, have kids, and get left by her. one can only hope...

afc

[Cubeboy]

Originally posted by Phinius
How high of a frequency a processor can achieve on a given process size has a lot to do with the amount of pipeline stages the core has. The 970 has 16 stages for integer-type instructions, 21 stages for floating-point instructions and up to 25 stages for Altivec (vector) instructions. Compare that to the Pentium 4 (20 pipeline stages) which will peak at about 3.2 GHz on the same process size as the 970 and it looks like it could very well be possible that the 970 will hit 2.6-2.8 GHz. If the person giving out this information to MacRumors is incorrect and just guessing, he or she is probably not off by much.

Look at it another way, the Power4 has 12 integer pipeline stages and the 970 has 16. The fastest POWER4 chip now runs at 1.7 GHz. Since the 970 has 1/3 more integer pipeline stages than the POWER4, then the 970 might be able to achieve 2.2+ GHz in the next few months. If this is in the realm of possiblity then I'd expect Apple to announce faster G5 speeds before the end of this year.

It isn't just pipeline depth that determines the clock rate of a processor, buffering, cache, die size, power dissipation, parallelism, latencies, components, and the general architecture of the CPU all go into determining how high a CPU can scale. We will use the Pentium 4 for example as it was designed to scale to high clock rates. A relatively small number of ALUs, FPUs, and General Decoders minimizes the amount of buffer and caches (which lowers clock speeds) needed to feed the processor. A tiny 8 kb data cache with a latency of 2 cycles along with a low latency L2 cache, and double pumped ALUs that lowers the latency of the ADD instruction, keeps latency low. Strong ALUs and Decoders keeps the distances between "pulses" (the point in time when registers can be loaded with new values) low, and finally we have the 20 stage pipeline enables us to do less per clockcycle and thus scale to much higher clock rates. Thats not to say the PPC970 can't scale any higher, I haven't looked over all the technical specifications yet, however, it's design philosiphy seems to be very different than that of the P4's with it's good number of ALUs/FPUs/Decoders, large L1 cache, and general focus on being able to execute multiple instructions every clock cycle.

[pimentoLoaf]

:(

I guess I'll have to wait 'til 2008 before getting a G5 with teeth -- I mean, why waste money on a 2ghz model when a 20ghz model is just around the corner 5 years from now?

[neutrino23]

I agree with both sides of this argument. We can already see today that in certain situations we already have computers that are fast enough. In a variety of business situations running FM Pro, a spreadsheet, a word processor, a browser and an email client then something like a 1GHz processor is plenty good. The market recognizes this and businesses are not upgrading as often as they used to.

On the other hand, new applications will arise that will take advantage of the much faster computers (and graphics chips). Many have been mentioned already.

AI will improve but personally I think they have a lot more theory to develop before they will benefit a lot from faster computers. Something key is missing at the core which is holding it back.

In many other areas though, scaling up the processing power (and memory, graphics chips, storage, broadband speed, etc.) will help a lot. Voice recognition will benefit. All sorts of modeling and rendering will become much more common. It remains to be seen whether this is done with style (like Keynote) or in a tacky fashion (like powerpoint).

We could forecast a revolution in computation, much as we had a revolution in electronics. Electronics replaced many mechanical things (springs, timers, indicators). Computation rides on electronics, but will replace the "analog" way of doing things.

For example, sensors in a room might be able to measure how a stereo is sounding and modify the speaksers/signal to produce really beautiful sound from relatively cheap speakers and amplifiers. Embedded computers in cars will continue to improve on the efficiency, smoothness and safety of driving. Embedded computers in displays will be able to model the behavior of nonlinear devices and adapt signals so that the produce undistorted results.

I have seen something like this already in the area of digital signal processors. In the old days we used analog amplifiers to shape and amplify pulses before they were digitized. Now the original signal is digitized and all of the pulse extraction, shaping and measurement is done numerically. As much as has been done it is just scratching the surface.

The hardware improvement will probably lead software development. Some of these are very hard problems. However, as a problem is solved it can be implemented very quickly and inexpensively.

As to how fast computer clocks can go, it is true that traditional semiconductor design seems to be coming up to physical limits. However, everyone is working on alternate technologies. I read somewhere that IBM forecast that somewhere around 2010 they would start introducing components using carbon nanotubes. These will be very fast.

[ralphh]

Originally posted by cuneglasus
This artcle is rubbish.It cant even get simple facts about the 970 straight.The 970 has 4 altivec pipelines not 2.(IT has a simple integer,complex integer,floating point and permute unit just like the G4).Why do people want to publish such lies.

Yeah, you're right about the four Altivec piplines. Whoever wrote this got that confused with the dual floating-point units in the G5. I'm hoping whoever this source is, he's merely lousy at passing info along and not a confabulator.

[Freg3000]

Originally posted by jouster
Hmmm. I dunno. I am embarrassed to admit that I don't know what serial or parallel ports are!!!!

Originally posted by XnavxeMiyyep
I dunno either. They're probably something already outdated. I think they still make PC's that use pre-USB ports to hook up the keyboard.

Don't be embarrassed, you are a better person for not knowing. :)

Yes, serial and parallel ports date back to the pre-USB days. The are antiquated and PC makes continue to put them on their machines. I think one of them is really big by the way, larger than an Apple ADC connector. Takes up more space on a laptop.

[Analog Kid]

Originally posted by Rincewind42
Coding in assembly is going the way of Latin. Being able to read assembly however is still the hallmark of a good programmer on his platform. There will always be a need to optimize your code because people will always expect things to be faster. So, you can calculate pi to 10 million digits in 2 seconds? Why couldn't you do it in 1 second while decrypting a 10 GB file? Sure, it's great to be able to put an app together without writing more than a handful of code - but if you can do it so can thousands of others and your app really isn't all that distinctive in that sense then is it? A distinctive app requires a programmer to take full advantage of all of the tools available to him or her - including optimizing.

Yup, and being able to read Latin is the sign of a good scholar, but very few people actually speak it... My metaphor is complete.:)

Why not be able to calculate pi to 10 million digits in half the time? Because very few people need to... There are a small number of people that need very high computation speed for special problems, but most of what I'm reading in this thread is from people that don't.

If you're doing genetics research, or particle physics then optimize your code. If you're building office apps and web browsers you want applications that add functionality and are able to interact cleanly. Making Safari an embeddable technology must surely bloat the code a bit, but rewriting and optimizing it for every new application just slows down development with no actual benefit to the user.

I'd guess that only 10% or less of Photoshop is hand tuned.

Office is one of the most successful applications around, and it certainly isn't optimized! Thousands of others could certainly build a word processor, but Word and WordPerfect took the top spots by adding features quickly.

I'm not contradicting you (is that the patter of thousands of little feet I hear?!), but trying to broaden the perspective. The initial voice recognition apps will be optimized for speed, the java app you send over an IM client to sing Happy Birthday to your girlfriend won't be.

[Analog Kid]

Originally posted by cuneglasus
This artcle is rubbish.It cant even get simple facts about the 970 straight.The 970 has 4 altivec pipelines not 2.(IT has a simple integer,complex integer,floating point and permute unit just like the G4).Why do people want to publish such lies.

4 execution units. 2 instruction queues.

http://www.arstechnica.com/cpu/02q2/ppc970/ppc970-2.html

Not saying the article isn't bogus, but that isn't the reason why...

[soggywulf]

Originally posted by solafide
Does anyone remember Apple's vision for computing in their short 1987 video "The Knowledge Navigator"? It included an intelligent agent that created a more natural interface. I can't wait to get rid of the keyboard!

Did anyone read "Snow Crash"? Screw google, I want that Librarian!

For anyone who thinks we don't need more computing power, sci-fi is a good place to start.

[elmimmo]

iMacs with decent speed are still a year off? I hope this does NOT confirm. pffff...

[soggywulf]

Originally posted by asim
what if you could hum a tune and it would automatically go through your itunes library and find that song, and if not find it from the itunes music store?

Now THAT is a kick ass idea! Here's something related that I've always thought would be incredibly cool. You know how there are some small parts of songs that sound really good to you, like they're hardwired to your soul? Such samples might come from a variety of music, and be unrelated to genre or band/composer. What if you could play those snippets, and your 50 THz machine would be able to search the "iTMS" to find all songs which have large portions of emotionally similar music?

[Dros]

Originally posted by Analog Kid


If you're doing genetics research, or particle physics then optimize your code.


What's great about future speed increases is that it will break down walls of what is possible for an average genetics or physics researcher. They will be able to write sloppy programs that aren't optimized and still get useful computation done. Right now too much of this is dependent on programmers who don't know the subject or researchers who can't code as well as the programmers. It'll bring supercomputer-level computation to the masses (if you count Ph.D. researchers as the masses).

Dros

[arn]

A little clarification added to the article

20-25GHz speeds are targeted by 2010-2011.

arn

[MacQuest]

Originally posted by skunk
Does 25GHz mean I'll be able to type faster?

:D ROTFLMAO!! :D

[Freg3000]

Originally posted by arn
A little clarification added to the article

20-25GHz speeds are targeted by 2010-2011.

arn

[begin sarcasm] OH! Well this isn't big news! I expected 25 GHz by 2011, but not by 2008! If it would have been huge if they got 25 GHz chips by 2008, but everyone will have 25 GHz by 2011. Stupid IBM. [end sarcasm]

Still good news, even if it is three years delayed. :)

[Phinius]

Originally posted by Cubeboy
It isn't just pipeline depth that determines the clock rate of a processor,

Yes, but if you increase the pipeline stages on a given design by 1/3 then it should top out at a much higher frequency. The POWER4 has 12 integer stages and the 970 has 16 stages. Yet the Power4 is at 1.7 GHz and the 970 is only at 2 GHz. IBM obviously increased the pipeline stages for the 970 to raise the topend frequency. How high of a frequency it can achieve is unknown at this time, but 2 GHz is very low for 16 pipeline stages. The AMD Athlon XP is at about 2 GHz and yet it only has 10-12 pipeline stages and as I said the 74557 G4 is expected to hit 1.8 GHz and it has a measly 7 pipeline stages.

We will use the Pentium 4 for example as it was designed to scale to high clock rates.

The 970 is also designed to achieve high clock rates. That's why IBM put 16 integer stages and 21 stages for floating point. You don't honestly believe that the IBM design is that horribly inefficient to not get significantly beyond 2 GHz on a .13-micron process do you?

Thats not to say the PPC970 can't scale any higher, however, it's design philosiphy seems to be very different than that of the P4's

The POWER architecture is designed to execute many instructions in every clock cycle, but with the added pipeline stages for the 970 it is also designed for much higher frequencies. That's obviously why IBM added the additional pipeline stages and a topend frequency of 2 GHz on a .13-micron process would be a patheticly poor result from adding all those extra pipeline stages.

[sedarby]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much? I guess that's what the iMac's and the like are for, though by then they will have 10 or 20 Ghz or whatever.

Remember this? :)

"640K ought to be enough for anybody"
- Satan, um, I mean, Bill Gates.

3 truisms of computers:
1). You can never have enough memory.
2). You can never have too much disk space.
3). You can never ever have a system that is too fast.

One use would be folding. Imagine how many units you could turn out with a system at 25 Ghz! Instead of a render farm you render on your home computer! Predicting weather and oil exploration will no longer be the private domain of super computers. The government will be changing their attitude when performance eclipses anything an Intel or compatible based PC can accomplish.

You get the idea. Too much speed is just not possible.

[Analog Kid]

Originally posted by Phinius
...a topend frequency of 2 GHz on a .13-micron process would be a patheticly poor result from adding all those extra pipeline stages.

If you're comparing to the Pentium, you're ignoring the difference in instruction set. Clock rate is determined not just by the number of pipe stages but also by how much work is done at each stage.

2GHz is already faster than the Power4 can clock, and it's the first "pressing". We'll probably see some speed bumps as IBM tunes their process and boosts their yields of the faster parts (didn't I hear rumors of a 2.8GHz part in testing?).

I'd guess we move to the 980 before the last pico-seconds are squeezed out of the 970 though. Then the 9x0 can keep abreast of the Power(x-3) developments.

Again referencing Ars Technica (without rereading the article so I may be wrong), their take was that the extra pipe-stages were inserted to allow the Alti-vec block to be bolted on to a remote area of the die. If I remember, they placed the extra stages in the decode, dispatch and completion phases of the pipe not in the execution phase...

[ZildjianKX]

Originally posted by soggywulf
Now THAT is a kick ass idea! Here's something related that I've always thought would be incredibly cool. You know how there are some small parts of songs that sound really good to you, like they're hardwired to your soul? Such samples might come from a variety of music, and be unrelated to genre or band/composer. What if you could play those snippets, and your 50 THz machine would be able to search the "iTMS" to find all songs which have large portions of emotionally similar music?

Actually the idea has already been done and researched by MIT I believe. A PC version was suppose to come out sometime this year, but who knows.

[MyLeftNut]

25ghz....seriously a practical application for even greater power will be the ships that will take us to Mars and beyond. Would you trust your life to a Wintel based ship? hehe

[Cubeboy]

Originally posted by Phinius
Yes, but if you increase the pipeline stages on a given design by 1/3 then it should top out at a much higher frequency. The POWER4 has 12 integer stages and the 970 has 16 stages. Yet the Power4 is at 1.7 GHz and the 970 is only at 2 GHz. IBM obviously increased the pipeline stages for the 970 to raise the topend frequency. How high of a frequency it can achieve is unknown at this time, but 2 GHz is very low for 16 pipeline stages. The AMD Athlon XP is at about 2 GHz and yet it only has 10-12 pipeline stages and as I said the 74557 G4 is expected to hit 1.8 GHz and it has a measly 7 pipeline stages.

The 970 is also designed to achieve high clock rates. That's why IBM put 16 integer stages and 21 stages for floating point. You don't honestly believe that the IBM design is that horribly inefficient to not get significantly beyond 2 GHz on a .13-micron process do you?

The POWER architecture is designed to execute many instructions in every clock cycle, but with the added pipeline stages for the 970 it is also designed for much higher frequencies. That's obviously why IBM added the additional pipeline stages and a topend frequency of 2 GHz on a .13-micron process would be a patheticly poor result from adding all those extra pipeline stages.

Your still using the most obvious (and most frequently used) answers (manufacturing capabilities, deeper pipeline) to address a question that delves far deeper than a simple explanation (deeper pipeline) can provide.

Let me ask you this question, a Pentium 2 has a 12 stage pipeline, a Alpha 21264 has a 7 stage pipeline, both cpus used .35 micron process and yet the Alpha was able to scale up to 600 MHz while the P2 was only able to scale to 300 MHz, how is this possible considering the Pentium 2's pipeline is nearly twice as long as the Alpha's?

Apparently, the answer lies beyond simply increasing the number of stages, but also to the design principles. Look back at my original list on all the aspects of a CPU that can affect clock rate, it turns out that our lovely Alpha engineers have covered nearly every one of those aspects, the cpu itself is highly paralleled and stages are kept extremely simple, this in turn, along with the Alpha chips strong ALU, and Decoders keep the distances betweens "pulses" extremely close. Cache and buffers will no doubt pose a problem as is true with all cpus of this class but apparently, it doesn't keep the Alpha from scaling. Latencies are kept at a minimal, ditto for power dissapation. No wonder the Alpha was able to scale twice as high as the Pentium 2, despite having a much shorter pipeline, nearly every aspect that affects scaling is far superior on the Alpha chip than the PII.

So how high would our Alpha cpu scale using the current .13 micron process? The general rule of process shrinks is a 50% increase in clock rate so moving from .35 micron process to .13 micron process (3 generations), our Alpha processor with it's "short" 7 stage pipeline will be at a amazing 2.025 GHz! Thats nearly as high as a Athlon which, with it's much longer 12 stage pipeline, has basically reached it's limit at 2.2 GHz and that's not assuming any improvements in Alpha architecture over all this time.

Apparently, as I have just proven with our Alpha example, shallower pipelines doesn't necessarily mean that a cpu can't scale to higher clock rates and deeper pipelines doesn't necessarily mean that a CPU can scale. It's all about design principles and expertise. Considering the Power4++ basically reached it's limit at 1.7 GHz, I wouldn't be surprised if a PPC970, with it's slightly longer pipeline would only be capable of scaling to around 2 GHz under the current micron process.

[BaghdadBob]

I'm with the gaming comment.

I belive OLEDs will finally yield VR goggles, and I'm looking forward to crapping my pants at hyper-realistic, finely tuned real-time physics involved, superior enemy AI utilizing Doom V in VR :eek:

Oh....yeah.....

Aside from that, the advances in technology and medicine that will come with farms utilizing that kind of power will be amazing.

The weather bit is BS, I don't care how much computational power you put forth, you cannot take all factors into account as is necessary to predict weather that accurately unless you're omnitient. Unless you don't believe that humans affect the weather, in which case you no longer have to be in the business of predicting human behaviour.

[Phinius]

Originally posted by Analog Kid
If you're comparing to the Pentium, you're ignoring the difference in instruction set. Clock rate is determined not just by the number of pipe stages but also by how much work is done at each stage.

2GHz is already faster than the Power4 can clock, and it's the first "pressing". We'll probably see some speed bumps as IBM tunes their process and boosts their yields of the faster parts (didn't I hear rumors of a 2.8GHz part in testing?).

IBM's German website already announced a upcoming blade server that will use 970 processors running from 1.8-2.5 GHz. So, the 970 will hit as least 2.5 GHz. It's also likely that a 2.1-2.5 GHz G5 would use 533 MHz DDR-II memory since Samsung already has it in production and Micron will manufacture it starting in late 2003. At 2.6 GHz the G5 will probably use 666 MHz DDR-II memory since 4X 666=2,664, which translates into 2.6 GHz processor using a 1.3 GHz bus speed.

Again referencing Ars Technica (without rereading the article so I may be wrong), their take was that the extra pipe-stages were inserted to allow the Alti-vec block to be bolted on to a remote area of the die. If I remember, they placed the extra stages in the decode, dispatch and completion phases of the pipe not in the execution phase...

If that is true then the POWER4+ processors should achieve greater than 1.7 GHz, because IBM already stated that the 970 will hit at least 2.5 GHz and the 970 is derived from the POWER4.

[Cubeboy]

Regarding "pulses":
As most of you have probably guessed by now, the distance between two pulse represents one clock cycle, therefore, with a given "length", more pulses will mean more clock cycles.

[Phinius]

Originally posted by Cubeboy
Let me ask you this question, a Pentium 2 has a 12 stage pipeline, a Alpha 21264 has a 7 stage pipeline, both cpus used .35 micron process and yet the Alpha was able to scale up to 600 MHz while the P2 was only able to scale to 300 MHz, how is this possible considering the Pentium 2's pipeline is nearly twice as long as the Alpha's?

The answer is that the Alpha 21264 and the Pentium 2 are completely different architectures. However, the POWER4 and 970 are not completely different chip architectures. The 970 does not have a large external L3 cache, unlike the POWER4 and the 970 has a L2 cache about a third the size of the POWER4. The 970 also does not have a chip interconnect fabric and an additional cpu and the POWER4 does. The 970 does have Altivec, but the core design of the 970 and POWER4 are basically the same.

IBM had made available a press release on their German website that stated a upcoming IBM blade server will use 1.8-2.5 GHz 970 processors. That officially confirms that the 970 will hit at least 2.5 GHz. An IBM spokesperson also stated that the 2.5 GHz speed will be on the .13-micron process.

Considering the Power4++ basically reached it's limit at 1.7 GHz, I wouldn't be surprised if a PPC970, with it's slightly longer pipeline would only be capable of 2-2.2 GHz.

Your assuming that the POWER4+ chip has topped out at 1.7 GHz, whereas IBM has made no such statement and your also stating that adding 1/3 more pipeline stages is 'slightly longer.' 1/3 is only slightly greater?

How can you make the claim that the 970 pipeline at 1/3 greater length than the POWER4 is only 'slightly' longer than the POWER4 pipeline? 1/3 greater seems to be a rather large step up in size to me, perhaps if Intel increases the 20 stage pipeline of the Pentium 4 by 1/3 and releases a new Pentium processor with 26 pipeline stages, then you will also state that it is has only a 'slightly' longer pipeline that the previous Pentium processor.

[Analog Kid]

Originally posted by Phinius
If that is true then the POWER4+ processors should achieve greater than 1.7 GHz, because IBM already stated that the 970 will hit at least 2.5 GHz and the 970 is derived from the POWER4. [/B]

Here's the Ars Technica link:
http://arstechnica.com/cpu/03q1/ppc970/ppc970-12.html

Note that they can't do much more than speculate either but it seems like they've put a lot of thought and research into it so I tend to trust their intuition.

Remember that the Power4 and 970 are being built on different processes. The Power4 is built on a more robust process for higher reliability, so even if it were identical to the 970 it would clock slower...

[soggywulf]

Originally posted by ZildjianKX
Actually the idea has already been done and researched by MIT I believe. A PC version was suppose to come out sometime this year, but who knows.

I'll believe that when I see it. ;) The marketing descriptions of some of these lab projects can grossly distort what they actually do.

[soggywulf]

Originally posted by sedarby
The government will be changing their attitude when performance eclipses anything an Intel or compatible based PC can accomplish.

Hmmm... Weren't we saying that about the 601? :)

[soggywulf]

Originally posted by asim
and some quality virtual-reality porn will require a quad GX (the "G ten") 42GHz with a 3d hologram projector... hopefully it will be ready by the time i find someone, get married, have kids, and get left by her. one can only hope...

LOL. Better skip the wife and kids if you want to afford the VR. :D

[soggywulf]

Originally posted by BaghdadBob
I belive OLEDs will finally yield VR goggles, and I'm looking forward to crapping my pants at hyper-realistic, finely tuned real-time physics involved, superior enemy AI utilizing Doom V in VR :eek:

Imagine hemishperical domes over each eye, to give you binocular depth perception along with peripheral vision. 3000x3000 res on each screen. Along with a lightweight gyro system to let the computer know which way your head is pointing. Now THAT would be a flight sim.

[Cubeboy]

Originally posted by Phinius
The answer is that the Alpha 21264 and the Pentium 2 are completely different architectures. However, the POWER4 and 970 are not completely different chip architectures. The 970 does not have a large external L3 cache, unlike the POWER4 and the 970 has a L2 cache about a third the size of the POWER4. The 970 also does not have a chip interconnect fabric and an additional cpu and the POWER4 does. The 970 does have Altivec, but the core design of the 970 and POWER4 are basically the same.

IBM had made available a press release on their German website that stated a upcoming IBM blade server will use 1.8-2.5 GHz 970 processors. That officially confirms that the 970 will hit at least 2.5 GHz. An IBM spokesperson also stated that the 2.5 GHz speed will be on the .13-micron process.

Your assuming that the POWER4+ chip has topped out at 1.7 GHz, whereas IBM has made no such statement and your also stating that adding 1/3 more pipeline stages is 'slightly longer.' 1/3 is only slightly greater?

How can you make the claim that the 970 pipeline at 1/3 greater length than the POWER4 is only 'slightly' longer than the POWER4 pipeline? 1/3 greater seems to be a rather large step up in size to me, perhaps if Intel increases the 20 stage pipeline of the Pentium 4 by 1/3 and releases a new Pentium processor with 26 pipeline stages, then you will also state that it is has only a 'slightly' longer pipeline that the previous Pentium processor.

You've just proved my point, the reason the Alpha scales higher is because they are different architectures in the sense that the way the Alpha was designed allowed it scale far higher than the Pentium 2 despite having a much shorter pipeline.

Apparently, having a longer pipeline which was the entire basis of your argument, doesn't mean that a processor will be able to scale higher nor does having a shorter pipeline mean a processor won't be able to scale higher. The entire point of my previous two post.

IBM has always released it's fastest Power4 possible and the current 1.7 GHz model has been around for quite a while. It's pretty reasonable to assume that it's the fastest Power4 possible with the current architecture and micron process.

The PPC970 has a 16 stage pipeline, the Power 4 has a 12 stage pipeline, putting this into context as well as considering what each stage actually does, I don't consider it very significant to scaling.

Steve Jobs has said that the PPC970 will scale to 3 GHz in 12 month(?) that makes sense considering moving to .09 micron process from the current .13 2 GHz will result in exactly 3 GHz.

So far, you've been unable to substantiate or prove any of your claims about pipelining and now your moving to articles. Do you care to provide a link?

[stingerman]

-The 970 is a superset of the Power4 core, not a subset. It contains additional instructions to allow for the quick migration of 32-bit OS's to 64-bits (namely, mapping virtual memory to physical memory, now that the processor can handle more RAM directly.) It contains an SIMD unit that Power4 does not. It can achieve higher frequencies. It contains a different Apple proprietary processor "coherent" interconnect technology.

- In a SMP environment the dual 970's cache technology behaves similarly to a dual core Power4. Namely, they combine each others L1 and L2 cache's, in effect equaling the size of the Power4 L1 & L2 Caches. (Opteron does the same thing.)

- The RAM speed does not delay the rate at which the Processor speed can be advanced. Right now, the RAM effectively runs at 800MHz and the 2GHz G5 at 1GHz. The bus is switched and gives every component a dedicated pipe. So the RAM can continue to run at 800Mhz while the G5's reach an FSB of 1.6GHz (3.2GHZ G5). Of course, it is always nicer to have ever faster RAM, but the speed of the system will still improve with a faster G5, especially when working with data in the L1 and L2 cache. Apple could also employ other techniques such as increasing the caches and adding superfast large L3 caches that would prefetch data from RAM. The L2 Catch also has a pre-fetch mechanism that reduces the effect of slower RAM.

- The 970 could easily go to 3.2 GHz and probably within 6 months, Steve was being conservative, can you blame him? They are already able to overclock some of the 970's coming off of the production line at 3.2GHZ. IBM always over-engineers.

- FP instructions have always been 64-bit, so OS X can already take advantage of the G5 64-bit FP units.

- Using the 970's new instructions for porting OS's to 64-bit, Panther will be able to easily address the entire 64-bit (42-bit address space) available to it. Jaguar 10.2.7 gives each processor 4GB each in a dual system, in the meantime.

- The 980's main design differences will be: 1. Shorter wire lengths within the core to increase speeds. 2. Hyperthreading (unlike Intel, the core can actually benefit from hyperthreading. 3. Better core power management, reducing power dynamically to less frequently used portions of the core. 4. .09 process, etc.

[Phinius]

Originally posted by Analog Kid
Here's the Ars Technica link:
http://arstechnica.com/cpu/03q1/ppc970/ppc970-12.html

Note that they can't do much more than speculate either but it seems like they've put a lot of thought and research into it so I tend to trust their intuition.

Remember that the Power4 and 970 are being built on different processes. The Power4 is built on a more robust process for higher reliability, so even if it were identical to the 970 it would clock slower...

According to a article written by David Wang at RealWorldTech.com, a IBM spokesperson informed him that the 970 and Power4 have the same thickness of traces. IBM did not use a more robust process for the Power4 than the 970. So, the argument for a slower or more reliable POWER4 than a 970 due to a different process is mute.

I would trust that the IBM German press release had accurate information and that the 970 will reach 2.5 GHz on a .13-micron process.

[RayCon]

In 1989, I bought a Mac IIcx. I was told it was state of the art. It ran at 16 MHz. I was told that if I was going to do graphics, I would need a bigger hard drive. Therefore, I bought a 40 MB.

In 1997, I bought a PowerMac 9600. It ran at 200 MHz (pre-G3). (16 to 200--or 12x-- in 8 years.) This time, it came with a 4 GB hard drive (10x).

In 2001, I bought a Quicksilver, operating with a blistering 867 MHz G4 processor. (>4x in 4 years or 54x in 12 years.) To boot, it came with a 60 GB hard drive. (15x larger in only 4 years, or a whopping 150x larger in 12!)

Oh, and after each purchase, I thought, "I'll NEVER fill THAT hard drive." Those who think that 10 or 12 or 25 GHz is too much will be amazed at the minimum requirements of the software when those machines are available.

[Phinius]

Originally posted by stingerman


- The 970 could easily go to 3.2 GHz and probably within 6 months, Steve was being conservative, can you blame him? They are already able to overclock some of the 970's coming off of the production line at 3.2GHZ. IBM always over-engineers.

That would mean the 970 would achieve the same topend frequency as the Pentium 4 on the .13-micron process size, which is highly unlikely.



- The 980's main design differences will be: 2. Hyperthreading (unlike Intel, the core can actually benefit from hyperthreading.

You mean simultaneous multithreading because Hyperthreading is Intel's trademarked name. IBM has stated that their version of simultaneous multithreading for the upcoming Power5 will perform like two processors going full throttle. As of yet Intel has only been able to achieve up to a 30% speed increase from using HyperThreading.

[Phinius]

Originally posted by Cubeboy

IBM has always released it's fastest Power4 possible and the current 1.7 GHz model has been around for quite a while. It's pretty reasonable to assume that it's the fastest Power4 possible with the current architecture and micron process.

The 1.7 GHz POWER4+ has only been out a few months and previous to that was a 1.4 GHz POWER4+ processor. So, IBM releases the fastest processor that they can at any given time, but that does not mean the it's the fastest POWER4+ possible with the current architecture and process size.

Steve Jobs has said that the PPC970 will scale to 3 GHz in 12 month(?) that makes sense considering moving to .09 micron process from the current .13 2 GHz will result in exactly 3 GHz.

The first 9XX processors made on a .09-micron process will almost certainly not be the fastest that IBM will eventually obtain on that process size. It takes time for manufacturing to increase the frequency of a processor on a given process size, it doesn't happen right from the get go.

An example of that would be Intel's upcoming Prescott Pentium processors which are expected to start at about 3.4 GHz and eventually it will peak at about 5 GHz over a year later.

Also the 970FX processor from IBM did not start out with a chip available at the topend speed. IBM moved it up to that speed about a year later.

Motorola is expected to have 7457 processors for sales this month and there will probably be availability for chips that run at the top speed of 1.3 GHz that is listed on Motorola's website. However, Motorola has been increasing the frequency beyond what's listed on the companies website for Apple's use. An internal Motorola document stated that the 7457 will peak at 1.8 GHz, yet don't expect that to appear immediately in a Apple computer, it will probably take some time for Motorola to get the chip up to that speed.

So far, you've been unable to substantiate or prove any of your claims about pipelining and now your moving to articles. Do you care to provide a link?

I cannot provide the link to the IBM German website press release about the 1.8-2.5 GHz 970 processors since IBM took the link down and it wouldn't do you any good if you don't understand written German.

[stingerman]

Originally posted by Phinius
That would mean the 970 would achieve the same topend frequency as the Pentium 4 on the .13-micron process size, which is highly unlikely.



[b]

You mean simultaneous multithreading because Hyperthreading is Intel's trademarked name. IBM has stated that their version of simultaneous multithreading for the upcoming Power5 will perform like two processors going full throttle. As of yet Intel has only been able to achieve up to a 30% speed increase from using HyperThreading.

Yes the 970 will achieve 3.2 easily before the 980 is released.

Intel's hyperthreading ideally is 30% faster but it also causes 30% speed reductions as well. It depends on the software. And most Wintel software does not behave well. That is why hyperthreading is usually turned off during benchmarks as it usually slows down the system. As far as its name, I prefer to use the same terminology to avoid confusion, the very thing the marketers try to create by renaming each others technologies. Actually an old IBM trick.

Hyperthreading type technology will work better on processors with a greater deal of instruction level parallelism (ILP). The whole purpose of hyperthreading is to convert threading into ILP. The theory is two threads can fill up the parallelism better than one thread. That is why the Intel P4 hyperthreading fails, the P4 is very low on the ILP, the very thing necessary for hyperthreading to work. SO instead of hyperthreading speeding up a P4, it creates another bottleneck in a lot of cases and slows down the whole system as one thread in a wait state hogs up a functional unit and all the instructions pile up behind it.

The Power4 and Power5 cores are ideally suited to hyperthreading will will in many cases double the number of instruction being processed in the same clock cycle!

[Wonder Boy]

[i]a 5 ghz pc/mac could very well be the last computer you ever need to buy. [/B]

thats what they said all those years ago about 60mgz and 120k computers...

[stingerman]

How about this for some speculation: The G5 case has also been designed to work with liquid cooling systems. Take a close look at the processor cooling fins. It appears that Apple already includes an intake and outtake tube as part of the design. Of course, I haven't seen one in person, only via analyzing the pictures. But it should be easiy to retrofit the very clean design with liquid cooling.

[ralphh]

Originally posted by stingerman

(snip)
Jaguar 10.2.7 gives each processor 4GB each in a dual system...
(snip)


Pardon my ignorance, can someone unconfuse me here? I thought that memory space was shared between processors in an SMP system. Otherwise, wouldn't two threads running within the same process have trouble sharing data if they were assigned different processors? Or is such an assignment not allowed?

[Analog Kid]

Originally posted by Cubeboy
Apparently, having a longer pipeline which was the entire basis of your argument, doesn't mean that a processor will be able to scale higher nor does having a shorter pipeline mean a processor won't be able to scale higher. The entire point of my previous two post.




I think his point is that given the same architecture, adding more (well designed and placed) pipe stages allows higher clock rates-- which is true.

Less logic between registers takes less time to propogate through. This is one way in which Intel has scaled the Pentium performance over the years.

If the 970 and the Power4 share the same basic architecture, which they do by all accounts, and the pipe stages are placed with the goal of increasing granularity in the critical timing paths (which I'm not sure they are) then you should be able to increase the clock rate of the chip.

This is not the only way to boost the clock rate, and it won't work without limitation, but staging the critical timing path does help.

Given what I've read about the architecture though, I don't know that boosting the clock rate over the Power4 was the reason for the added stages. It would seem like a lot of trouble to go through if your design goal was to save effort on the architecture development...

Based on where the stages were added, it looks as though placing the Alti-vec block in a sub-optimal location on the die actually added a timing constraint in communicating the data to and from the main execution core. The stages may not have been added to boost the clock rate, but rather to keep the stages balanced so the clock rate didn't have to decrease...

[tazznb]

Originally posted by Veldek
I'm quite sure that I read some days ago, that IBM has a 970 working at 3.2GHz dissipating 82W, which is as much as the Intel chips at the same frequency. So it doesn't seem impossible for them to reach the 3Ghz barrier with a 970.

Anyway, as IBM is working on the Power5 (the 980 will be a derivative of it, the G4 has nothing to do with either of them) and on the Power5+, I think we may expect a lot in the near future.

The P4 @ 3.2Ghz dissipates up to 100W, and can be used as a second oven.

[rjstanford]

Originally posted by ralphh
Pardon my ignorance, can someone unconfuse me here? I thought that memory space was shared between processors in an SMP system. Otherwise, wouldn't two threads running within the same process have trouble sharing data if they were assigned different processors? Or is such an assignment not allowed? The original quote should have referred to each process being allocated 4gb, not each processor. Unless I am very much mistaken, which is always a possibility.

-Richard

[Analog Kid]

Originally posted by Phinius
According to a article written by David Wang at RealWorldTech.com, a IBM spokesperson informed him that the 970 and Power4 have the same thickness of traces. IBM did not use a more robust process for the Power4 than the 970. So, the argument for a slower or more reliable POWER4 than a 970 due to a different process is mute.

"Although the PowerPC 970 is a part that would cost considerably less to manufacture and sell, its performance actually exceeds the POWER4 processor in many areas. The reason for this apparent paradox is that the POWER4 processor had been designed for the high cost, continuous availability server market, and in some areas, performance had been traded off to obtain near-absolute reliability guarantees. As an example, in an article published in Microprocessor Reports in 1999, IBM described its use of thicker gate oxides in the POWER4 processor to obtain a failure rate that is two orders of magnitude better than comparable processors from most other manufacturers. The cost of the thicker oxides is the reduced drive current of the transistor and consequently slower switching speed of the transistors on the POWER4 processor. In the case of the PowerPC 970, the processor does not need to meet similar reliability requirements as the POWER4 processor, and as a consequence, circuit and process technology can be tweaked to obtain higher performance by trading away the near-absolute reliability required by the POWER4 processor."
--David Wang
RealWorldTech

http://www.realworldtech.com/page.cfm?AID=RWT101502203725

I don't have time to go through all the links... Can you link an article with updated information?

Thanks

[Analog Kid]

Originally posted by tazznb
The P4 @ 3.2Ghz dissipates up to 100W, and can be used as a second oven.

That's just amazing... The heat dissipation guys really are the unsung heroes of the computing world.

I seem to remember my sister having a Betty Crocker oven of some sort that used a lightbulb as a heat source... I'm sure it wasn't a 100W bulb.

[Haberdasher]

Originally posted by soggywulf
Now THAT is a kick ass idea! Here's something related that I've always thought would be incredibly cool. You know how there are some small parts of songs that sound really good to you, like they're hardwired to your soul? Such samples might come from a variety of music, and be unrelated to genre or band/composer. What if you could play those snippets, and your 50 THz machine would be able to search the "iTMS" to find all songs which have large portions of emotionally similar music?

There's already a service for this for cell-phones. Dial the service, put the phone to a song on the radio, or you can whistle/hum the tune. The service searches its database by matching certain peaks and troughs (I think), and returns with a possible match. Apparently it's pretty accurate, and is getting better all the time (able to ignore more and more background noise, softer music, etc.)

Sorry I can't remember the name, but it IS out there (it was on CNN or I read it in Time...) Try looking for it, I remember it sounded wicked sweet. :D

[beloit08]

I noticed that the article says the G5 is coming into the PB relatively soon. Anyone have a clue as to how they're going to cool the thing? I've heard they already get pretty hot with the G4 inside. Putting a G5 in there that runs at 97w (or whatever)--we'll all end up with 17" CookBooks. Or something cooled with liquid nitrogen...

[XForge]

Originally posted by Shadow_Raptor
As for everyone worrying about the 20-25Ghz 45nm process rest assured that this problem has been expected for some time now. It is true that the theoritical limit for current processors is up around 30Ghz. After that everything gets too small, too fast, too power hungry and too hot.

Okay, liquid-nitrogen-cooled Power Macs!!! Woo!!

[edit] I had no idea beloit08 had posted that about Powerbooks already... jinx!!

[ralphh]

Originally posted by Analog Kid

I seem to remember my sister having a Betty Crocker oven of some sort that used a lightbulb as a heat source... I'm sure it wasn't a 100W bulb.

The Easy Bake Oven!!:D

Yeah, two 100-watt bulbs supplied all heat, one above and one below with a couple of simple right-angle reflectors focused on the cake pan. It really cooked! My sister made a lot of great stuff with hers, including a small but very good apple pie I loved.

Just replace a P4's heat sink with a reflector and rig a sliding tray for the little cake pan and you're all set! ;)

[XForge]

Originally posted by CmdrLaForge
I don't know as well. I even don't know for what I could need a G5. At the moment I have a iBook with 900MHz. It does all I want at a speed that is quite enough for me.

One should note that the US version of Macworld magazine was cryptically saying, in a review of the 900MHz G3 iBook, that G3s may have "compatibility issues" in the future. Due to their publishing cycle they do know stuff ahead of time, and are limited only by their nondisclosure agreement. Wildassed guessing I'd say whatever comes after Panther won't boot on G3s. Wildassed guessing mind you.

[metamotion]

All this speculation about IBM's hardware prowess (and thus Apple's) is very interesting, but I submit that the new architecture carries great significance for those of us who care about Apple's survival and the direction of the PC industry.

I suspect that Mr. Steve is now poised to ramp up Marklar, not put it on the back burner. Here's why: Release of an Intel/AMD compatible operating system would have been suicide for the lucrative hardware operation before now. Some intrepid Wintel users may have been willing to spend $129 for an elegant, crash-resistant alternative to the dark side, but widespread acceptance and real profits will take several years to build, even with the amazing success of the Windows Ipod that we will see this holiday season and next year.

At the same time, a Marklar OS blazing away on a 3.6 Ghz P4 box would have snuffed sales of Apple hardware within five minutes. But now it appears likely that IBM/Apple has nearly equalled the top of the wheezing Wintel combine, and will pull off a minor leapfrog in about 8-12 months.

When G5 boxes appear that can run Panther dramatically faster than Longhorn or Marklar on a top-end Dell, look for Mr. Steve to release the thing with a bet-the-farm marketing campaign. Complete with optimized Adobe products and a serious Office competitor from Apple that combines Keynote, a revamped Appleworks and perhaps even iLife on Intel/AMD boxes. I think that July 2004, in time for Christmas season, is within reach for such a move.

Having, at long last, some speedy architecture means that Apple can launch Marklar and still sustain its outrageous profit margins on boxes --- and still build market share among professional, scientific and the vanity consumer markets that will pay for great industrial design and blazing speed. Why not buy an Apple, if it is demonstrably faster than an Intel/AMD box running the same OS?

For the multitudes, the vast unwashed who don't need more speed until Internet connections get much faster, those lucky folks get to see a real alternative to the Windows monopoly. An operating system that really works. Bundled with applications that work, not just exist to crush competitors and compel bug-laden upgrades. The next Windows XP upgrade will have serious competition. Sadly, any dreams of Linux making it in the consumer/business space will be toast.

The end game for all of this is that Apple has another chance to do what it failed to do in the mid-1980s: make great software for a living, compete with Microsoft and return billions to its shareholders. And let Sony, Dell or Whoever Is Next kill themselves trying to make money building consumer hardware.

As for me, I can't wait to buy that $800 system running an Apple OS with a 19 inch LCD and plenty of expansion ports. And real competition will mean that my art department can make that spinning beachball go away without bankrupting the company.

[Frohickey]

PowerPC, the architecture is derived from IBM's POWER architecture (Performance Optimized With Enhanced Risc). Apple approached IBM and got Motorola into the partnership to make PowerPC chips. It borrowed heavily from the POWER architecture.

The first ones were the PPC601, then the PPC603, and PPC603e. Then the PPC604 came along, which allowed multi-processor configurations. There is a PPC604e, IIRC. There is a bastardized one from Exponential, which never did pan out, but that was just a process change instead of any chip architectural improvements.

After this, the PPC750 came out, which is also known as the G3. There are still variants of these out since its a pretty good low power and decent performer. Along this time, the PPC7400 was in the works and its claim to fame is the addition of vector instructions to the PowerPC architecture. It was around this time that Amelio tried his hardest to crashdive Apple to the ground. At least, he made a good decision to bring el Stevo back. :)

This crash-diving made IBM antsy, and so they decommitted from the PPC7400 (G4). This left Motorola doing most of the development and bug fixes. When the iMac using the G3 took off, el Stevo was able to convince IBM back into the fold, but they were now behind. Since they decommitted, Motorola was not going to let IBM back in so easily. Here is where the bad feelings came in.

When the G4 came out, Apple convinced Motorola and IBM to play nice, and there were a few agreements signed, and IBM was back making PPC, this time the PPC7400 variant. Some of these actually came out into Macs.

IBM wanted back in, but Motorola likes being the only supplier of chips, so this left IBM playing second-fiddle. What they needed was a way to get back into the leap-frogging development style, since following Motorola's lead means that they will always be behind Motorola.

While all of this is happening, Motorola came out with the PPC7410, which was supposed to give the G4s a shot in the arm as far as clock speeds, but this was not to be. It did give a little lower power, at the same speeds, same power, at slightly higher speeds, but it wasn't the substantial clock speed increase. But it so happens that the PPC7450 was just around the corner anyway.

This was to be a 4 core processor. Similar to what the Power4 is, only with vector instructions that IBM thinks only makes a chip larger and more complicated. But the 4 core version was not meant to be. It was too expensive, and not suitable for placement into the inexpensive iMac and iBook. So, it had to go on a diet. First, it became a 2 core processor, and finally, became the single core processor that we know is the PPC7450.

The PPC7450 didn't get a substantial increase in speed either. It was larger, with 7 stage pipes, but it topped out at 733MHz. The PPC7400 with its 4 stage at 500MHz (PPC7410 at 550MHz).

So, as usual, the next chip is coming from Motorola, the PPC7455. This is the PPC7450 with a new process and a few tweaks, not a new architecture. So, this is where things sat until IBM finished working on the PPC970.

And this is where we are at now.

As to what the future will bring? I'm not supposed to say or else the people in the black helicopters are going to get me. :p

*shhhh*
What's that?
*CRASH*
NOoooooooooo! I swear I didn't say anything!
I know my rights, you can't make me do anything!
GET YER HANDS OFF OF ME, YA FILTHY APES!!!

[MarkCollette]

Originally posted by Wonder Boy
thats what they said all those years ago about 60mgz and 120k computers...

I think that we do have to acknowledge that there are different markets for computing power, and that while we do know that there will always be a market for the utmost processing power, that general consumers way well cease to be a part of that market.

I think that general users may have an interest in real-time encoding of DVDs, which will lead to further processing demands when we go to HD-DVD, but I don't see the media companies going beyond HDTV for several decades. Tthink how many decades it's taken for HDTV to come, and it's not even pervasive yet. So that means we'll need more than 5X the computing power, that we currently have, available at consumer price points, maning we will need almost 10X the computing power available to the consumer, prosumer and workstation markets. That would mean running at 20GHz, or dual 10GHz, or quad 5Ghz. That would mean, according to Moore's Law, we would reach a significant consumer drop-off point between 2005-2008. Since IBM might not keep up with Moore's Law for the processors it provides to Apple, that schedule might slip to 2009.

Try to remember I'm only talking about consumers, not professional. I expect a combination of a continuous drop-off in demand, with a refocussing on making parts cheaper, more reliable, more portable, and better able to interface with external systems. These external systems might include faster storage, new wireless technologies, or even things like direct access to our brains. Whatever. That might push the schedule for a few more years. None-the-less, by 2015 the consumer market will no longer be pushing innovation, and thus innovation will receive less financing, and so will slow. Further advancement will return to the area of big companies, militaries, etc.

I know based this all on video encoding, but I believe that that one demand is representative of consumer demand for CPU processing, in general. Ie, a game or word-processor need not be that fast, in general.

Just look at other industries, like flight, where early developments were involved with the military, then the consumer market pushed developments for a while, and then the majority of development returned to the folds of the military, once consumer cost/benefit objectives had been met. Of course we'll probably all be flying supersonic ten years from now, but that just supports my point that progress continues, it just doesn't trickle down quickly to the consumer area, after consumer markets aren't the primary focus.

This is all very good for Apple, since they already focus on other issues than raw speed. One might argue that once computers can process everything we'd need them to, then maybe we'll have a chance of them beginning to do what we actually need them to do - suck less.

[Cappy]

I just want one fast enough to take over the world. I don't think that is too much to ask for, no? It's not like I'm asking for a tank of sharks with laserbeams on their heads. ::sigh::

[stingerman]

Originally posted by Frohickey
PowerPC, the architecture is derived from IBM's POWER architecture (Performance Optimized With Enhanced Risc). Apple approached IBM and got Motorola into the partnership to make PowerPC chips. It borrowed heavily from the POWER architecture.



Nice history. However, I don't like the expression that PowerPC was derived from Power. The AIM group completed the Architecture and they selected parts from the Power platform to help expedite the actual first PowerPC implementation which found it's way first in the Power2 processor. A subset of Power turns out to be an even further subset of the PowerPC architecture, and then IBM initially added only the 32-bit parts of the PowerPC Architecture (which was actually 64-bit architecture on paper.) They basically took a mustang stripped it to its frame kept some of the parts and rebuilt a Corvette on top of it.

Neither Power nor Power2 were 64-bit processors though PowerPC specs were. And Power is not binary compatible with PowerPC, Power2 is however. The Power3 finally implemented in 1996 the full 64-bit architecture and was binary compatible with the Power2 and any other PowerPC processor. The Power3 went through three planned major enhancements. Then the Power4 added dual cores, higher parallelism, superscalar clocking and a very advanced processor interconnect technology that allowed IBM to include 8 processors in one Power4 module.

The Power5 will add a hyperthreading type capability allowing each core to come very close to doubling its processing capability. The Power5 will also feature a redesigned shorter wiring between parts and better internal organization. It will also implement advanced power management that will allow it to turn off and reduce power to unused or infrequently used processor sections. An even higher degree of parallelism, larger caches, and a .09 process.

The 980 (G6) will be a further superset of a single 980 core, with an enhanced and better integrated Altivec unit, as well as Apple's Proprietary Processor Interconnect technology.

[stingerman]

Steve Jobs promised 3 GHz within 12 Months, that allows for less than 12 months. This will have to happen with the G5 (970). Expect the G5 to hit 2.4 GHz by October. Apple needs to erase the past image due to the Motorola crap and the only way to do this is to beat their targets and continually surprise the markplace. They need to prove that the G5 was not a fluke but a real change in direction and accelleration. This is the only way business will commit large investments in the G5 platform, they need to have the comfort that the platform will not die on the vine.

We will probably see a new single processor line released this fall for the rest of us as the PowerMac G5 is really a Power User/workstation class machine that will eventually only be dual processing powerhouses. (Take a look at the Apple Store, everyone is mainly buying the dual G5's.)

The single processor G5 PC will be geared for the corporate/enterprise and small business office productivity user. Display sold seperately. Apple simply does not have a computer for this target audience. The eMac for education and budget PC, the iMac for the consumer market, the PowerMac for the Power User / Workstation market. It's a big hole Apple needs to fill and the reason why they do not have general corporate marketshare, they currently simply do not sell a product for that space. With all the changes in Panther to appeal to the corporate types, a Mac must be in the wings that complements Panther. And some office productivity software to reduce the cost of switching, since no one wants to pay yet again for Office.

[Phinius]

Originally posted by Analog Kid

I don't have time to go through all the links... Can you link an article with updated information?

Thanks

Unfortunately I could not find the statement. But as I recall David Wang stated in Arstechnica or at Realworldtech.com that a IBM engineer informed him that the traces were left the same on the POWER4 and 970 because IBM could find only a neglible performance boost by changing it. So David stated that he was incorrect in assuming that IBM reduced the RAS for the 970 by making the wiring thinner.

[jettredmont]

Originally posted by daRAT
Ahh yes he did, do a web search for that phrase. All the famous quote sites list him as the author. He made it in 1981.

And at that time, it WAS enough :P

Yes, but they aren't primary sources. Show me a primary source. The only primary source for that quote that is well known is Mr Gates himself, and he flatly denies ever having said such a thing (in fact, makes a fairly convincing argument that at least between MS and IBM said quite the opposite, and that 640k was a compromise between IBM's proposed 512k and what he wanted, which was presumably much larger).

No matter how many times someone is misquoted, it is still a misquote ...

[GregA]

Originally posted by metamotion
I suspect that Mr. Steve is now poised to ramp up Marklar, not put it on the back burner. Here's why: Release of an Intel/AMD compatible operating system would have been suicide for the lucrative hardware operation before now.<snip>

Having, at long last, some speedy architecture means that Apple can launch Marklar and still sustain its outrageous profit margins on boxes<snip>. Why not buy an Apple, if it is demonstrably faster than an Intel/AMD box running the same OS?Most people seem to say Apple has huge margins on their hardware - yet they seem to assume a $0 cost for MacOS X? Shouldn't we assume the OS cost is AT LEAST the upgrade cost of $129? Any $800 PC with Marklar would have to make Apple at least that ($200?).

Maybe Steve will consider selling OSX on IBM's blades? or for AMD's 64bit chips. It gives Apple a tamer entry to 'alternate' hardware, and still differentiates them as a 64bit OS. AMDs chip may emulate some PPC code easier too.0

I've given up on Marklar - much as I like the idea of Apple everywhere (even an MS Office alternative on Marklar AND Windows). I'm still considering trying to load Darwin on an old PentiumII I have lying around... but I might just Redhat it.

[eric67]

Originally posted by stingerman
Yes the 970 will achieve 3.2 easily before the 980 is released.

Intel's hyperthreading ideally is 30% faster but it also causes 30% speed reductions as well. It depends on the software. And most Wintel software does not behave well. That is why hyperthreading is usually turned off during benchmarks as it usually slows down the system. As far as its name, I prefer to use the same terminology to avoid confusion, the very thing the marketers try to create by renaming each others technologies. Actually an old IBM trick.


The Power4 and Power5 cores are ideally suited to hyperthreading will will in many cases double the number of instruction being processed in the same clock cycle!

you are wrong and right,

wrong about the gain due to HT, actually it has been widely proven on pc web sites that HT enable processor vs HT disable processor have a gain around few % (5-7%) maximum obtained for specific application : +15%. but this when applications have been written to take advantage of the HT, when it is not the case, actually HT leads to a decrease of the performance...

you are right the Power4 and 5 and derivates are ideally suited for HT, but it might require to have applications written specifically for HT.

[MarkCollette]

Originally posted by stingerman
...
We will probably see a new single processor line released this fall for the rest of us as the PowerMac G5 is really a Power User/workstation class machine that will eventually only be dual processing powerhouses. (Take a look at the Apple Store, everyone is mainly buying the dual G5's.)
[/B]

One of the main strengths of the G5 is its ability to do SMP better then P4s. Anyone who's carefully read over the benchmarks would be cheating themselves to buy a single 970 system. Well, except for those who just can't afford them :rolleyes:

Anyways, I agree with everything you've said, but I think you've forgotten about the 1.4 GHz chips that we haven't seen in use yet. These might never be seen in an Apple (ala 68060), or they could be the basis of an iMac or an Xserve revamp. The Xserves ( 1U ) probably don't have the leisure of cooling space that the PowerMacs have, yet have even more need (being servers) of the 64bit address space (or 40 some odd bits as of now). In many instances it is better for a server to have more, cheaper, cooler CPUs than a single, faster, hotter CPU. This is the whole reason why IBM is making the 970s, for blades.

I would expect a use of the 1.4 GHz chip in the next few months, perhapse in a redo of the previous strategy of releasing the OS X Server software before the regular OS X, since there is a lesser focus on the GUI in the server products? By this I mean an Xserve with Panther Server, before September, when the regular Panther release will occur.

[stingerman]

Originally posted by eric67
you are wrong and right,

wrong about the gain due to HT, actually it has been widely proven on pc web sites that HT enable processor vs HT disable processor have a gain around few % (5-7%) maximum obtained for specific application : +15%. but this when applications have been written to take advantage of the HT, when it is not the case, actually HT leads to a decrease of the performance...

you are right the Power4 and 5 and derivates are ideally suited for HT, but it might require to have applications written specifically for HT.

I was being generous, but I agree with your %, I mentioned that 30% was an ideal number, should have been clearer.

HT on 980 will not require special coding. The OS sees it as dual processors and will give each virtual processor its own thread. This will work great with Cocoa apps which naturally lend themselves to being very well threaded. OS X itself is designed as a highly threaded OS and will take great advantage of the multi-threading for all its work. Couple this with Quartz Extreme for handling many of the visual aspects and we got ourselvse one highly parallel processing machine and OS.

[jettredmont]

Originally posted by Rincewind42
Sure, it's great to be able to put an app together without writing more than a handful of code - but if you can do it so can thousands of others and your app really isn't all that distinctive in that sense then is it? A distinctive app requires a programmer to take full advantage of all of the tools available to him or her - including optimizing.


I disagree. A distinctive app requires more than programming prowess. It requires intelligent design, from the UI to the inner workings.

Programming bottlenecks in assembler can often buy you 10-20% performance improvement overall. If you're crunching numbers, that's awesome. But then again, if you're crunching numbers and you put the same amount of effort into improving the overall algorithms of your application at a higher level, you might find yourself gaining 50%-100% performance.

If your app isn't about raw number-crunching, then the user is far more likely to benefit from an intuitive interface than even a 20% performance gain. 20% sounds like a huge performance gain, and it looks great on a bar chart, but the simple fact rules performance optimizations:

Users generally do not notice less than a doubling of performance (ie, 100% gain).

So, yes, when you have the most perfect algorithms and you have the most perfect UI then it is time to move on to bottleneck-busting. It is even time to bust bottlenecks when you can't think of any more fundamental improvements. But if you buy that 20% improvemtn by moving everything to hand-tuned and hard-to-maintain assembly then you've traded a vast amount of fundamental future improvements for a single immediate improvement.

Which brings us back to the question: what does 25GHz buy us? It doesn't buy us Photoshop rendering the same old scenes twenty times as fast. It doesn't buy us MS Word being able to spell-check in a hundredth of a second instead of a fifth of a second.

Greater hardware buys the ability to take applications to the next level. It allows developers to think in larger terms, to avoid having to worry (as much) about the nigley details. It allows the use of garbage collection schemes that work instead of pedantic allocation-deallocation rules and constraints. It allows reuse of code in ways the original developer may well have never foreseen and certainly not optimized for.

Or it just makes Quake run really really fast and get like 20,000 frames per second.

[h'biki]

Originally posted by Analog Kid
Yup, and being able to read Latin is the sign of a good scholar, but very few people actually speak it... My metaphor is complete.:)

And its a curious fact that the decline of latin has gone hand in hand with the rise of mass literacy.

Similarly, the decline of assembly as a programming language has gone hand in hand with the democratisation of computing and programming.

I don't think its necessarily an indicator of the quality of someone's code if they can program in assembler. Good code is good problem solving. You could recode into assembler and make the thing execute faster but your code can still be ineffecient.

If you want to continue with your language analogy, knowing latin doesn't guarantee that you're going to be a good writer.

[eric67]

Originally posted by stingerman
I was being generous, but I agree with your %, I mentioned that 30% was an ideal number, should have been clearer.

HT on 980 will not require special coding. The OS sees it as dual processors and will give each virtual processor its own thread. This will work great with Cocoa apps which naturally lend themselves to being very well threaded. OS X itself is designed as a highly threaded OS and will take great advantage of the multi-threading for all its work. Couple this with Quartz Extreme for handling many of the visual aspects and we got ourselvse one highly parallel processing machine and OS.

I agree with you, but I still think that all the G5 revision and follow up (based on 980 and behond) should be dual processor machine by default, even if in the futur we will have HT....

[mproud]

Remember when manufacturers would boast how fast their CDs could read? 2x CD-ROM, 4x CD-ROM! Some games would actually require a quad speed CD-ROM to take advantage of it...

...and so today, about 10 years later... does anyone really care anymore how fast a CD reads? Perhaps writing and rewriting - but, that'll be irrelevant soon too - you'll see...

But what I'm trying to get at is, there will be more technologies that will take advantage of new stuff, but only so much. At some point, gradually, the clock speed will matter less (home users in mind anyway).

Although companies will continue to play up the numbers for as long as they can, the focus will generally shift towards features, ease of use, experience, support, price, and overall value - stuff that truly plays a little bigger part today than people let it have - and also qualities that Apple currently excels at.

Just like CD reading speeds or general HD seek time have become unimportant, the same could go for any speed indicator on computers - or at least when 25 GHz computers come out (or whenever cinematic-quality graphics can render near-instaneously).

[mproud]

Originally posted by MarkCollette
I think that we do have to acknowledge that there are different markets for computing power, and that while we do know that there will always be a market for the utmost processing power, that general consumers way well cease to be a part of that market.
...
This is all very good for Apple, since they already focus on other issues than raw speed. One might argue that once computers can process everything we'd need them to, then maybe we'll have a chance of them beginning to do what we actually need them to do - suck less.

...and then I read your post.

Exactly :)

[soggywulf]

Originally posted by MarkCollette
I know based this all on video encoding, but I believe that that one demand is representative of consumer demand for CPU processing, in general. Ie, a game or word-processor need not be that fast, in general.

Don't you believe it. Games will always, *always* need more power. It doesn't matter how much you give it, it will not be enough. 25 GHz will not be enough. Games are limited only by imagination, and imagination is an incomprehensibly enormous thing--larger than the universe itself.

I see the requirements for games accelerating far ahead of video work, at least for some subsystems of the computer.

Also, remember that the games out-revenued Hollywood last year (I think). The game industry is a major driver of innovation, and I see that as only increasing in the future.

[soggywulf]

Originally posted by GregAussie
Most people seem to say Apple has huge margins on their hardware - yet they seem to assume a $0 cost for MacOS X? Shouldn't we assume the OS cost is AT LEAST the upgrade cost of $129? Any $800 PC with Marklar would have to make Apple at least that ($200?).

This argument makes no sense. Windows is a crappy OS, but it costs money to make too. There must be a ton of cost involved in making XP run on such a diverse variety of hardware. Yet you can buy extremely cheap and fast machines that run WinXP.

Now if you want to argue economies of scale, that would be legitimate.

[rocks1985]

This is My First Post on the MacRumors Forums!!! :)

[h'biki]

Originally posted by GregAussie
Maybe Steve will consider selling OSX on IBM's blades? or for AMD's 64bit chips. It gives Apple a tamer entry to 'alternate' hardware, and still differentiates them as a 64bit OS. AMDs chip may emulate some PPC code easier too.0


Actually, my money would suggest that Steve's 'options' may actually mean an OS X running on Power4 and Power5s - possibly badged as an XServe.

The ultimate in badass servers.

(And they get to surprise everyone when they announce their next gen XServers which everyone expects to be 'just' G5s. Nup. Apple is going balls-to-the-wall into the server market.)

I don't think I'm totally dreaming either - yes, I know the current iteration of XServes is a 1U cae which doesn't have the cooling requirements of the 970... but we'll see.

[GregA]

Originally posted by soggywulf
This argument makes no sense. Windows is a crappy OS, but it costs money to make too. There must be a ton of cost involved in making XP run on such a diverse variety of hardware. Yet you can buy extremely cheap and fast machines that run WinXP.

Now if you want to argue economies of scale, that would be legitimate. But that's exactly what I mean. When Dell calculates their profit on sale - it includes having paid something for the OS. From what I have read, Apple considers the cost of their OS on each machine to be $0. Thus if they 'profit' $400 on a given machine, maybe the OS should be considered to cost $200 of that. So their profit on each computer is not as high as people thought.

The economies of scale are also related - maybe the cost of maintaining an OS is 5 times higher for Apple than Microsoft. So $40 per PC for Microsoft might be realistic and $200 per Mac for Apple (until they start selling 10 times as much).

Maybe I've just misread - and when Apple claims a high profit on their machines they've already factored OS cost???

[soggywulf]

Originally posted by Haberdasher
There's already a service for this for cell-phones. Dial the service, put the phone to a song on the radio, or you can whistle/hum the tune. The service searches its database by matching certain peaks and troughs (I think), and returns with a possible match. Apparently it's pretty accurate, and is getting better all the time (able to ignore more and more background noise, softer music, etc.)

Sorry I can't remember the name, but it IS out there (it was on CNN or I read it in Time...) Try looking for it, I remember it sounded wicked sweet. :D

Wow, sounds cool! I'll have to check that out. But, I have my doubts. Matching songs I can almost believe, but finding "emotionally similar music"--I seriously doubt it. If something like that really exists, it would seem to me that this system would be mind-bogglingly more complex and advanced than anything else out there today, both algorithmically and computationally. Again, I'll have to hear it to believe it.

Anyway, thx for the pointer! :)

[sigamy]

How many fans will that 20Ghz Power Mac have???

[soggywulf]

Originally posted by GregAussie
But that's exactly what I mean. When Dell calculates their profit on sale - it includes having paid something for the OS. From what I have read, Apple considers the cost of their OS on each machine to be $0. Thus if they 'profit' $400 on a given machine, maybe the OS should be considered to cost $200 of that. So their profit on each computer is not as high as people thought.

My point is that this does not explain price/perf differences. Both platforms have an OS and hardware, which both presumably have similar development costs on either platform. Thus, "The OS costs money" is a misleading statement. The OS costs the same for both sides.

What make the difference are (1) economies of scale, as you mention later, (2) lack of a vertical monopoly on the PC side, and (3) OSX is better than XP, so Apple can charge extra for that.

Edit: re-reading your post, I can see that you were saying the same thing...I guess we pretty much agree. :)

[soggywulf]

Originally posted by sigamy
How many fans will that 20Ghz Power Mac have???

None. It will dissipate 1.5 watts.

[Hattig]

Originally posted by daRAT
Ahh yes he did, do a web search for that phrase. All the famous quote sites list him as the author. He made it in 1981.

And at that time, it WAS enough :P

No, he didn't make the quote. It is simply one of the widely spread myths. At worst, it was a comment that was taken out of context ("[some people think that] 640kb is enough [but I don't]")

[patrick0brien]

Originally posted by C14ru5
What's the point of having eLiza error correction in a consumer processor? I thought error correction only was necessary in server processors like Power4 and Power5. I mean, how often do calculation errors occur in a processor chip?

-C14ru5

Well, the article says eLiza would be used to primarily step in when the branch prediction unit fails.

"when the branch prediction unit fails" - That sounds bad. what they're trying to say is "when the branch prediction unit doesn't predict the righ branch". It is prediction after all, and like Joe Weatherman, can't always see the future perfectly. :D

eLiza is then used to alleviate the data traffic pileup that would occur "when the branch prediction unit fails".

[Docrjm]

Originally posted by sigamy
How many fans will that 20Ghz Power Mac have???

Much more than the 3% marketshare and hardly any inside the box!:D

[rog]

Originally posted by XForge
One should note that the US version of Macworld magazine was cryptically saying, in a review of the 900MHz G3 iBook, that G3s may have "compatibility issues" in the future. Due to their publishing cycle they do know stuff ahead of time, and are limited only by their nondisclosure agreement. Wildassed guessing I'd say whatever comes after Panther won't boot on G3s. Wildassed guessing mind you.

Well I don't think Macworld has any more insight than any of us on this forum. They've long since ceased to be anything but a cheerleader for Apple. Apple could take 5 years to bump the iBook to 950 MHz and their cover would boast "FASTEST IBOOK EVER!!!!" and they'd go on and on about it and take off 1/2 a mouse for some random reason.

Of course Apple will drop G3 support before G4 support because OSX is increasingly making use of altivec.

[NNO-Stephen]

Originally posted by sigamy
How many fans will that 20Ghz Power Mac have???

several million. you know how us Mac nuts love the "next big thing" :)

[JoeRadar]

Originally posted by arn
This could be true... it could be made up. I don't know.

The real unknown was Apple switching from Motorola to IBM and the PowerX family. Once that was revealed, searching for information on Apple’s future path is not that difficult -- just go to google. Suggested searches:

power5 linux boot

power4 power5

power4 power5 power6

power4 power5 site:ibm.com

[JoeRadar]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane

See The Age of Spiritual Machines: When Computers Exceed Human Intelligence. (http://www.amazon.com/exec/obidos/tg/detail/-/0140282025/qid=1057596108/sr=1-1/ref=sr_1_1/104-9505976-2159138?v=glance&s=books) Kurzweil estimates a $1000 computer will reach human brain computation rates in the 2020s and will reach the computation rate of the entire population by the 2050s. Lots of interesting ideas.

In Taking the Red Pill (http://www.amazon.com/exec/obidos/ASIN/1932100024/qid%3D1057596850/sr%3D11-1/ref%3Dsr%5F11%5F1/104-9505976-2159138), a series of essay inspired by the Matrix, Yale philosopher Nick Bostrom (http://www.nickbostrom.com/) posits that we may already just be simulation programs that think we are real.

And of course, there is always HAL from 2001 (http://www.amazon.com/exec/obidos/ASIN/0451457994/qid=1057597009/sr=2-1/ref=sr_2_1/104-9505976-2159138) and Star Trek's holodeck. (http://www.amazon.com/exec/obidos/tg/detail/-/0262631873/qid=1057597145/sr=1-1/ref=sr_1_1/104-9505976-2159138?v=glance&s=books)

[PowerBook User]

It's nice to see that IBM is already working on several new processors. This can only be good for Apple. I am a bit disappointed that the iMacs might not have a G5 for another year. That could leave a huge performance gap between the iMac and the Power Macs.

[wms121]

this is more poop for those who enjoy good speculation.

The URL below is for the Japanese division that IBM contracts to:

http://www.research.ibm.com/thinkresearch/pages/2000/200011_optical.shtml


There may be no "50GHz" pure semi chips past 2012..an experimental "POWER12" is supposed to be running in a 'wormhole generator' out at Site4 (Area51). It has 4 POWER6's running on top of a SiON optical chip RAM in a MCM type direct custom prototype. There are several "teraflops on a chip" type proto's in IBM's research labs, most use optical memory.

4 of these would make a nice "Mac Cube"..and would have optical power and telecom (Infiniband 32x/64x) networking.

<==embedded snitch in DoD land

[MacManDan]

Regarding: humming a tune and iTunes will find the song and play.

Originally posted by ZildjianKX
Actually the idea has already been done and researched by MIT I believe. A PC version was suppose to come out sometime this year, but who knows.

This is true .. MIT has a working prototype (that works VERY well). MIT's AI Labs also have a great deal done on many things I've seen on this post .. visual/voice recognition has had years and years of development. One of MIT's AI Lab's robots had "eyes" and learned to play with blocks. That same robot could track and follow someone waving a ball in front of it. This was all done with a couple of networked PIIIs .. it's ridiculous to think you need a 25ghz computer to do that.

True .. the technology does need to be improved to even have a prayer of reaching a human's visual/audible recognition abilities, but some of the stuff that the AI Lab comes up with is just amazing. Since I go to school at MIT, I have seen a lot of this stuff first-hand. If you're ever in the Boston area, go to the MIT Museum in Cambridge .. you will see you don't need a 25ghz machine (or even a 5 ghz machine) to do very good recognition and AI.

25ghz machines can do a LOT more! Think bigger! Think WeatherPop v2000 with a huge prediction model that could accurately forecast the weather for your front yard for the next month. THAT is more along the lines of what a 25ghz machine could do.

[iLilana]

Originally posted by Ja Di ksw
Not to be rude, but what does someone need 25 Ghz for? Honestly, that is just insane, especially if they are dual processors. I know people making music or animation or whatever would like them, but for your average Joe, do we really need that much? I guess that's what the iMac's and the like are for, though by then they will have 10 or 20 Ghz or whatever.

Remember this? :)

"640K ought to be enough for anybody"
- Satan, um, I mean, Bill Gates.

simulated holographic display technology

[-hh]

Originally posted by macrumors12345
Wow, Geordi, you are the first person I have seen who is planning to buy the 1.8 Ghz model. I have seen a number of posts saying "I plan to get the Dual 2 Ghz G5," but this is the first one I've seen saying "I want to get the Single 1.8 Ghz model." Good to know that someone is buying them. Still haven't seen anyone going for the 1.6 Ghz.


Well, you can put me down on the list of "lesser" PowerMacs too.


I don't need the pricetag of a DP 2GHz, so I'm debating between a 1.6 and a 1.8GHz.

In running some numbers, it costs $275 to bring the $1999 1.6Ghz G5 up to be "equal" with the 1.8GHz G5 for the RAM & Hard Drive. This means that if you want those options anyway, the price premium for the extra 200MHz is effectively only $150 ($2,249 vs $2399), which is probably worth buying (an extra +12% in clock speed for only +6% cost).

However, the caveat is the usual one ... this is paying Apple's RAM prices. The $150 effective price differential jumps from $150 to $250 when 3rd Party RAM is used to expand to 512MB.

Nevertheless, I'm leaning towards the 1.8 because it uses the same RAM as the 2GHz and the 1.6 doesn't...I have the "bad feeling" that this might affect the machine's lifecycle 3-4 years from now in terms of what gets support dropped first.

Overall, if I could get a DP 1.6GHz for $2300, I might consider it. And a DP system should be theoretically avaiable as an aftermarket retrofit, if Apple really wants to permit it.

Finally, a similar cost analysis can be done with the DP 2GHz: Apple offers a BTO option on the 1.25GHz G4 to upgrade it from SP to DP for $300 which can be used to SWAG the SP->DP cost. The markup from the 1.8Ghz is $600, and when you value the +200MHz bump at $150, the videocard at $50 and the 2nd chip at $300, you come up with $500 worth of "value". When viewed in this context, the price points seem very well worked out.


-hh

[XForge]

Originally posted by Jeff Harrell
AI is not a computationally bound problem. It's not like we have AI systems out there right now that can do the job but are too slow. AI is a theoretically bound problem: we don't even know if it's possible yet, much less if it's possible with existing technology, much less whether existing technology is fast enough.

OTOH artificial stupidity is easy as pie. See for example, PowerPoint. (grin)

[pixelpusher]

One of the fundamental truths about computing is that if you build it, they will come.

Finding applications of massive supercomputing power is easy. Just see what applications are running only on the fastest supercomputers. Weather prediction, etc...

How about human genome stuff. I met a geneticist that thought trying to do the Celera route of brute force decoding the human genome would never work because of how long it would take. He didn't take into account moore's law. He was working on finding a more elegant solution when the microprocessor let other guys just do a brute force solution.

A lot of hard problems are gonna be solved with brute force and superfast mircoprocessors.

On another note, I can imagine cell phones that can scan a potential mate in a bar and determine if their genetics are good for breeding.

Hopefully sooner, I'd like to attach some 'trodes on my head and get the machine to Image what's in my head so I don't have to work an physical input device. Photoshop with a stylus go away; After Effects rendering begone!

[kenaustus]

The Apple-IBM relationship is exciting because it is allowing Steve J to start taking on the Wintel world in a very powerful way. The new G5's are basically the "entry level" computers for this round of development and you can be sure that the folks at Apple are working hard to ensure that all hell breaks loose over the next year or two.

While rapid increases in speed will be great for power users the increase will also be important in the competitive environment - who will be king of the hill? I for one would love to see a Mac that with a faster Ghz rating that the top of the line Pentium - it might shut up some of the bashers from the Wintel world for a while!

The challenge today is one of engineering - e.g., how do you bring the new chips to an efficient cooling level in a PB? That will probably have to wait until the 980 is released with the 0.09 micron process. Once there, however, one might see an optional dual processor PB in the 15" and the 17". Just the thought makes it hard to sleep at night.

There is also going to be significant enhancements in the consumer lines and I would guess that the Mhz gap may be closed at the various pricepoints in that market, depending on what IBM can come up with in the way of processors.

Apple is going to continue to pour money into OS X and the Mac apps, with probably one or more major upgrades to OS X before Longhorn is finally released. (This is one of the reasons why I didn't hesitate to pay for 10.2 and will be first in line for 10.3 - I want to see OS X increase their lead over WIndows.) The introduction of iTunes & the Music Store under Windows is also going to generate significant revenues for Apple and, for the first time, let a lot of Wintel users see what the Mac world is really about - something that will motivate them to look at Macs when they need or want a new computer.

I believe that these factors are moving to a point where the lower end Macs will be able to generate a significant increase in market share, which provides the economies of scale needed to bring prices down, or performance significantly up (or both) for the entire range. Gonna cause a lot of worry at MS!

We are still at the point where the G5 has been announced, but not delivered, and are talking about how soon the G5 will double in speed, how soon we'll hit 20 Ghz and, in another thread, who long it will be before Apple releases the 30" Cinema display. Damn, it is an exciting time to be a Mac user & lover!

[MacManDan]

Originally posted by pixelpusher
On another note, I can imagine cell phones that can scan a potential mate in a bar and determine if their genetics are good for breeding.

Hopefully sooner, I'd like to attach some 'trodes on my head and get the machine to Image what's in my head so I don't have to work an physical input device. Photoshop with a stylus go away; After Effects rendering begone!

Now THAT's thinking big. It's stuff like what pixelpusher is talking about that may seem like science fiction but have more credibility when talking about technology in 10 years. In 10 years, with the release of this 25ghz processor and other technology that we can only dream of, the machines we are working on now (even those blazing new G5s) will seem like primitive stone-age tools. It's the ideas that people like pixelpusher have (rather than used-up, already-done ideas like voice/face recognition etc) that keep R&D people on their toes and come up with stuff that will make use of the brand-new technology of the time. Just think .. 10 years ago, how many of you could have imagined dual 2.0ghz G5 processors that can render life-like 3d images, edit movies in professional quality, etc etc?

[jettredmont]

Originally posted by -hh
Well, you can put me down on the list of "lesser" PowerMacs too.


I don't need the pricetag of a DP 2GHz, so I'm debating between a 1.6 and a 1.8GHz.

In running some numbers, it costs $275 to bring the $1999 1.6Ghz G5 up to be "equal" with the 1.8GHz G5 for the RAM & Hard Drive. This means that if you want those options anyway, the price premium for the extra 200MHz is effectively only $150 ($2,249 vs $2399), which is probably worth buying (an extra +12% in clock speed for only +6% cost).

However, the caveat is the usual one ... this is paying Apple's RAM prices. The $150 effective price differential jumps from $150 to $250 when 3rd Party RAM is used to expand to 512MB.


Are you sure about this?

Note that 256MB RAm from Apple comes as 2x128 and 512MB RAM from Apple comes as 2x256 (everything's 2x for memory on the new systems because of the interleaved design). Granted, you have four memory slots, so with 2x128 you can buy two more 128 sticks to bring it up to 512MB overall, but there is a significant advantage to having 2.256 instead of 4x128 ... And I SERIOUSLY doubt you'll find 512MB of PC2700 RAM for $25 ($125 - claimed $100 savings in going to third-party RAM) unless you know of some deal that I don't see out there (Crucial is good memory and they have 2x128 at $48 and 2x256 at $82 ... maybe you can find someone to buy your excised 2x128 to offset the cost of the 2x256 ...)

Plus, as you said, this is PC2700 RAM on the 1.6GHz and PC3200 RAM on the 1.8GHz, which pretty much accounts for the cost differential you see between the two machines, even were the processors to be the same ... I think the 1.8GHz machine is, overall, a better deal than the 1.6GHz machine (and the 2x2GHz step-up is even more appealing).

Yes, if you don't need/want the power, save your money ... but in the overall scheme of things it is fairly rare for the high-end step-ups to be as attractively priced as these are ...

[eric_n_dfw]

...everyone keeps talking about the PM G5's 9 fans and how:
a) The Xserve's 1U box won't be able to keep a G5 cool
b) the PB would fry your lap off with a G5
c) The extra fans are because the G5 runs so hot.

Wrong - Wrong and Wrong.

From ArsTechnica: http://www.arstechnica.com/cpu/02q2/ppc970/ppc970-1.html
As you can see from the table, the 970 at 1.8 GHz is much closer to the G4e than to the P4 2.8 GHz in terms of power dissipation. This means that Apple will be able to use this chip in the kinds of innovative enclosure designs that make their hardware continually appealing, regardless of how it performs. Furthermore, a 1U, 970-based version of the XServe is not out of the question. And if you consider the fact that the 970's power consumption at 1.2GHz is a mere 19W, it's almost certain that we'll see a future notebook from Apple based on the new chip.

I predict dual 2Ghz G5 XServe's very soon as well as 1.2 (maybe 1.6) Ghz G5 PB's early next year.

Will the G5 XServes be loud? Heck yes - but who cares?

[g3ski]

I am always waiting for the next thing. now that I know the 9900 will be 20+Ghz, i will have to wait unil 2010 to upgrade

:confused:

[BaghdadBob]

You know, when we get to these speeds we're going to see amazing things going on in entertainment. Think "mass proliferation of real time 3D" and multiple times more realistic surface rendering in movie-grade 3D -- resulting in indistinguishably real animated characters (unlike the Phantom Menace).

And "Final Fantasy the Movie VIII: More Inconstruable Flashbacks and Indescernable Personal Problems"

[soggywulf]

Originally posted by MacManDan
Just think .. 10 years ago, how many of you could have imagined dual 2.0ghz G5 processors that can render life-like 3d images, edit movies in professional quality, etc etc?

Hmm...10 years ago DOOM came out. I think just about everyone on the planet was imagining life-like 3D in the future. Not that we're there yet with the G5, by any means. I think a lot of these ideas that you consider old-hat are not developed anywhere near their potential, or not really developed at all. A PIII robot following a ball != Terminator. :)

[jettredmont]

Originally posted by eric_n_dfw
...everyone keeps talking about the PM G5's 9 fans and how:
a) The Xserve's 1U box won't be able to keep a G5 cool
b) the PB would fry your lap off with a G5
c) The extra fans are because the G5 runs so hot.

Wrong - Wrong and Wrong.


You are quoting an ArsTechnica article which was based off of preliminary specs put out by IBM. We now have the real specs, and the heat dissipation picture is far worse for the 970 at 1.8GHz than previously believed.

You will not find the 970 in laptops anytime soon according to some quite reliable Apple honchos ... 'course they could be aiming to deceive, but ...

[daveL]

Originally posted by jettredmont
You are quoting an ArsTechnica article which was based off of preliminary specs put out by IBM. We now have the real specs, and the heat dissipation picture is far worse for the 970 at 1.8GHz than previously believed.

You will not find the 970 in laptops anytime soon according to some quite reliable Apple honchos ... 'course they could be aiming to deceive, but ...
Most of the posts here seem to equate 90 nm to the 980. That's so, but there's no reason why the 970 won't move to that process, and it could very well come before the 980. It will be a lot easier for IBM to move the existing 970 to the new process than bringing out the 980. With the 980, you have a new design *and* a new process, while the 970 will be a proven design. I guess my point is that we may get a mobile 970 out of the 90 nm process. Moving to the 90 nm process can boost the speed and/or reduce the power, take your pick in terms of the tradeoffs. So, maybe we get a 3GHz+ PM 970 and a 2+GHz PB 970, or something like that, from the 90 nm process.

Just a thought.

[Analog Kid]

Originally posted by g3ski
I am always waiting for the next thing. now that I know the 9900 will be 20+Ghz, i will have to wait unil 2010 to upgrade

:confused:

It's worse than that. By 2010 you'll be reading rumors of 200GHz SiGe machines just around the corner and you'll have to wait until 2015 when you'll hear about...

;)

[wizard]

Hi guys;

Another way to look at this is that the 1.6 serves the needs of people who need backwards compatability with PCI. Thinking this way places the 1.8 at the begining of the modeline, or entry level as far as the new archetecture goes. So you there is more value in the 1.8 realtive to the 1.6 then first appears. This is why I consdier the 1.6 to be grossly overpriced, it is a machine designed to provide backward compatability as far as I can see and really should not be priced as close as it is to the 1.8.

Dave


Originally posted by jettredmont
Are you sure about this?

Note that 256MB RAm from Apple comes as 2x128 and 512MB RAM from Apple comes as 2x256 (everything's 2x for memory on the new systems because of the interleaved design). Granted, you have four memory slots, so with 2x128 you can buy two more 128 sticks to bring it up to 512MB overall, but there is a significant advantage to having 2.256 instead of 4x128 ... And I SERIOUSLY doubt you'll find 512MB of PC2700 RAM for $25 ($125 - claimed $100 savings in going to third-party RAM) unless you know of some deal that I don't see out there (Crucial is good memory and they have 2x128 at $48 and 2x256 at $82 ... maybe you can find someone to buy your excised 2x128 to offset the cost of the 2x256 ...)

Plus, as you said, this is PC2700 RAM on the 1.6GHz and PC3200 RAM on the 1.8GHz, which pretty much accounts for the cost differential you see between the two machines, even were the processors to be the same ... I think the 1.8GHz machine is, overall, a better deal than the 1.6GHz machine (and the 2x2GHz step-up is even more appealing).

Yes, if you don't need/want the power, save your money ... but in the overall scheme of things it is fairly rare for the high-end step-ups to be as attractively priced as these are ...

[soggywulf]

Originally posted by wizard
Another way to look at this is that the 1.6 serves the needs of people who need backwards compatability with PCI.

Don't think so, Dave. The PCI-X slots in the higher machines are already backwards compatible with current PCI cards, except maybe for some ancient ones. I don't think there are many of those 5V cards in use. Perhaps really high end stuff like Media 100 etc? But for those folks, clearly they would be getting new cards with their 2 gig duals (or the other way around, 2 gig duals with their new cards).

However, I am also curious as to why they decided to design a different motherboard for the 1.6. It must have cost them some significant money, just for a market differentiator. Seems to me that money could have been better used for other things, like maybe dropping the price points a bit. :D

[cgc]

Originally posted by QCassidy352
good lord, makes me not even want a g5 anymore! ;) I hope it's all true. That sounds great.

One question though- can someone explain "power 4" and "power 5" as compared to 970 and 980? The 970 is a power 4 derivative, while the 980 is a single core power 5? Is the G4 also a power 4 derivative? Is the 990 also a power 5 derivative? Thanks for the help.

The G4 is a 604E derivative and the G3 is a 603 derivative (and lacking decent FPU, if I remember).

[MarkCollette]

Originally posted by soggywulf
However, I am also curious as to why they decided to design a different motherboard for the 1.6. It must have cost them some significant money, just for a market differentiator. Seems to me that money could have been better used for other things, like maybe dropping the price points a bit. :D

Maybe the 1.6 motherboard was an initial attempt, like a prototypig effort, and the 1.8 and 2.0 motherboard was a later improvement, after they had better access to DDR400 memory?

[XForge]

Originally posted by soggywulf
Did anyone read "Snow Crash"? Screw google, I want that Librarian!

For anyone who thinks we don't need more computing power, sci-fi is a good place to start.

I wanna be able to Google from inside my head like Mona Lisa Overdrive.

[Rincewind42]

Originally posted by jettredmont
I disagree. A distinctive app requires more than programming prowess. It requires intelligent design, from the UI to the inner workings.

Which is all in a programmer's toolbox. Programming is more than just writing code (or putting together already created modules). There maybe others dictating what goes into your application, but it is up to programmers to make it work.

Programming bottlenecks in assembler can often buy you 10-20% performance improvement overall. If you're crunching numbers, that's awesome. But then again, if you're crunching numbers and you put the same amount of effort into improving the overall algorithms of your application at a higher level, you might find yourself gaining 50%-100% performance.

I specifically stated that writing assembly is going the way of the dodo (or rather, latin). That doesn't mean that you don't need to be able to read it. If you want to understand what the compiler is emmiting and how to get it to do better in a higher level language you need to know as much as you can about your architecture.

And to optimize you must know that you are using the best algorithm. If you aren't already using the best algorithm, then you are wasting your time optimizing suboptimal code.

If your app isn't about raw number-crunching, then the user is far more likely to benefit from an intuitive interface than even a 20% performance gain.

Agreed. But a fast and intuitive interface is even better.

20% sounds like a huge performance gain, and it looks great on a bar chart, but the simple fact rules performance optimizations:

Users generally do not notice less than a doubling of performance (ie, 100% gain).

Users may not notice a 20% increase in short tasks, but they notice it in long tasks. And while 20% in one place may not be all that great, 20% in a lot of places can make quite a big difference.

If Word takes 20% less time to spell check my project report, then my compiler doing a make on the latest version of my software will go that much faster. So an optimization in a short (but often repeated) task will make a long process go faster. And while it isn't noticed by the user, it's appreciated.

So, yes, when you have the most perfect algorithms and you have the most perfect UI then it is time to move on to bottleneck-busting. It is even time to bust bottlenecks when you can't think of any more fundamental improvements. But if you buy that 20% improvemtn by moving everything to hand-tuned and hard-to-maintain assembly then you've traded a vast amount of fundamental future improvements for a single immediate improvement.

No one ever said to do everything in assembly for speed - and in fact my post said the opposite. And I never said that you optimize everything - optimizing too much is just as bad as doing no optimization work at all. And optimizing too early is the root of all evil.

And you will never be done if you are waiting for the perfect algorithm & interface because the perfect interface & algorithm have something in common - the work is done before you even ask :D.

Greater hardware buys the ability to take applications to the next level. It allows developers to think in larger terms, to avoid having to worry (as much) about the nigley details. It allows the use of garbage collection schemes that work instead of pedantic allocation-deallocation rules and constraints. It allows reuse of code in ways the original developer may well have never foreseen and certainly not optimized for.

This line has been repeated ad nauseum for over 10 years. So why hasn't anything changed? Because with faster processors users expect faster applications that do more. They don't care if you the programmer have to deal with pedantic allocation rules & constraints or how much code you get to reuse - they want their work done fast, now and if they shell out money for a faster computer they expect it to go faster.

And regardless, newer computers/cpus don't always make software go excessively faster. Just look at the P4 when it came out - many many benchmarks shows that for a lot of tasks the P4 was slower than a P3. Whose to say that this rumored 9900 won't run code slower than the 990 that it replaces? If you want to take full advantage of the 9900 you would have to optimize for it. Sure if you wait long enough things would just be faster but do your customer really want to spend more money for the same speed as before?

Which brings us to the next question - what about the people running on slower machines? They want to use your application too, and they want to not feel as if they need to buy a new computer to use it. Unoptimized programs may run fast enough on your development machine, but often your development machine is also a high end machine. Your users often will have weaker hardware than you. If speed is only acceptable on your development machine, then it will almost certainly be unaccetable on a low end machine.

For the record, much of this same conversation took place on one of Apple development lists last week. That discusion concluded that optimization is a necessary step in development, not just so that things work well on older hardware, but also so that they work well on newer hardware.

[Analog Kid]

Maybe I'll be surprised and we'll get a G5 Powerbook tomorrow, but people sound like they're going to riot without one...

The PBG4 is a beautiful machine, and it's one of the few areas where Apple has been competitive with Wintel in performance.

Originally posted by eric_n_dfw
...everyone keeps talking about the PM G5's 9 fans and how:
a) The Xserve's 1U box won't be able to keep a G5 cool
b) the PB would fry your lap off with a G5
c) The extra fans are because the G5 runs so hot.

Wrong - Wrong and Wrong.

I predict dual 2Ghz G5 XServe's very soon as well as 1.2 (maybe 1.6) Ghz G5 PB's early next year.


Everybody keeps pointing to the power draw of the 970 and then talk about how low power the system is...

Nobody talks about the power of driving a 64bit bus at 1GHz.

Nobody talks about the power of the off chip memory controller and system ASIC ("one of the fastest ASICs in the world").

Nobody talks about the power of a 128bit wide, double pumped, 400MHz RAM system.

No way I'm putting one of these on my lap until the CPU, the bridge ASIC and quite possibly the memory go through process shrinks.

Given my 'druthers, I'd integrate the memory controller onto the CPU too to remove the I/O power of one bus.

The next PBs are going to be based on cooler G4 systems. If the dual processor PB turns out to be anything it'll be 2 G4 cores on a single die...

That's the only thing that makes sense to me. Putting a G5 into that enclosure now would certainly make waves in the industrial design world when the unit glows a really funky red color during operation, but I don't think it's practical.

2 CPU modules would take an absurd amount of power when you could use a dual core chip.

No I don't know anything ya'll don't, I'm just making sense of what I see...

Is the G4 memory controller on chip?

[Rincewind42]

Originally posted by jettredmont
You are quoting an ArsTechnica article which was based off of preliminary specs put out by IBM. We now have the real specs, and the heat dissipation picture is far worse for the 970 at 1.8GHz than previously believed.

You will not find the 970 in laptops anytime soon according to some quite reliable Apple honchos ... 'course they could be aiming to deceive, but ...

Even after seeing the real specs I don't think that the dissipation picture is far worse. Yes the case has 9 fans. But first off realize that the case was designed for cooling 2 970s at 2Ghz. That's nearly 100 watts by itself (assuming IBM did hit their estimated mark). It is also cooling up to 8 GB of RAM, which isn't exactly cool running either. Then there is the DVD-R drive, the 2 up to 250 GB SATA hard drives, and 3 PCI-X slots (that can provide more power and run much faster than the PCI slots in the previous generation). And finally, there is the Mirrored Door G4s. You know, the ones that are known colloqually as 'Wind Tunnels'. I'm willing to bet that on top of the actual cooling concerns was the concern of what these machines would sound like. There are 9 fans in the case so that they are Quiet (and they are). I can easily imagine concerns of these new machines being nick-named "Hurricane" or "Tornado" G5s.

Now as for when they make it into a PowerBook, that is totally up to Apple. I still think that if they wanted to right now they could put a 1.2 Ghz G5 into a PowerBook. They'd probably fly off the shelves and not be all that hotter than the 1Ghz G4s (if at all). But they could have decided that the G4 has enough life left in it to go another generation in the PowerBooks. Or they could have discovered that the system controllers would consume too much power. Or any number of other things could have caused Apple to hold off on G5 PowerBooks. But I also think that most of these reasons will evaporate when the G5 goes to a 90nm process. Unfortunately, as far as the rumor scene knows, that won't happen until early next year or so.

[-hh]

Originally posted by jettredmont
Are you sure about this?

Note that 256MB Ram from Apple comes as 2x128 and 512MB RAM from Apple comes as 2x256 (everything's 2x for memory on the new systems because of the interleaved design). Granted, you have four memory slots, so with 2x128 you can buy two more 128 sticks to bring it up to 512MB overall, but there is a significant advantage to having 2.256 instead of 4x128 ...

And I SERIOUSLY doubt you'll find 512MB of PC2700 RAM for $25 ($125 - claimed $100 savings in going to third-party RAM) unless you know of some deal that I don't see out there (Crucial is good memory and they have 2x128 at $48 and 2x256 at $82 ...

Oops - my math was slightly off - I misrecalled Apple's RAM markup as $150 ... you're right its only $125. In any event, while Mac compatibility needs to be verified, Pricewatch currently lists:

$59 - PC2700 DDR 512MB
$27 - PC2700 DDR 256MB
$21 - PC2700 DDR 128MB

So the net savings would be $125 - (2*$21) = $83 instead.
IMO, that's enough of a savings to be worth making the effort.



...maybe you can find someone to buy your excised 2x128 to offset the cost of the 2x256 ...)


So long as there's "empty" slots (meaning either literally empty, or full but with no need for a slot to add more RAM), the old RAM is not valuable enough to be worth pulling.

And based on the above prices, I'd not bother with the 128's...the 256 is the current sweet spot ("knee in the curve"), so I'd buy two of those instead. In conjunction with leaving the 2x128's in the machine, I'd end up with 768MB total.


-hh

[eric_n_dfw]

Originally posted by Analog Kid
Everybody keeps pointing to the power draw of the 970 and then talk about how low power the system is...

Nobody talks about the power of driving a 64bit bus at 1GHz.

Nobody talks about the power of the off chip memory controller and system ASIC ("one of the fastest ASICs in the world").

Nobody talks about the power of a 128bit wide, double pumped, 400MHz RAM system.

No way I'm putting one of these on my lap until the CPU, the bridge ASIC and quite possibly the memory go through process shrinks.

Given my 'druthers, I'd integrate the memory controller onto the CPU too to remove the I/O power of one bus.

The next PBs are going to be based on cooler G4 systems. If the dual processor PB turns out to be anything it'll be 2 G4 cores on a single die...

That's the only thing that makes sense to me. Putting a G5 into that enclosure now would certainly make waves in the industrial design world when the unit glows a really funky red color during operation, but I don't think it's practical.

2 CPU modules would take an absurd amount of power when you could use a dual core chip.

No I don't know anything ya'll don't, I'm just making sense of what I see...

Is the G4 memory controller on chip? very good points - maybe the PB G5 is farther off than I initially predicted. As far as the actual wattage of the G5's, can anyone point me to an official (Apple or IBM) page that lists them? All I can find is some article that states the 2Ghz G5 is a whopping 97 Watts. I find it a little hard to believe that numbers would have jumped that high from IBM's original predictions for the 1.6 - but I'm no electrical engineer so what do I know?

I still think we'll be seing G5 XServes - just with very high RPM fans.

[Phinius]

Originally posted by stingerman
Steve Jobs promised 3 GHz within 12 Months, that allows for less than 12 months. This will have to happen with the G5 (970).

If the G5 is moved to a smaller process within a year, IBM's transition to a smaller process will still effectively be at least 9 months behind Intel's move. I'd still expect to see the G5 at 3 GHz on a smaller process size and not at that frequency on the same process size as is being used now. IBM has already admited that they are working on the next generation of the G5 which makes it a strong possiblity that the G5 could be moved to a smaller process size within a year. Afterall, IBM does have a new state-of-the-art fabrication facility and the companies next generation process size is ready to go.

Where did you get 3.2 GHz for the G5 on the same process size? It looks like you just pulled that out of thin air or is it because the Pentium 4 is now at 3.2 GHz and you feel that IBM/Apple has to match that?

The G5 could be bumped up to higher frequencies if IBM increases the volts which is what Motorola did to increase the frequency of the G4 for Apple. However going up to 3.2 GHz from the 1.6-2.0 GHz that the G5 is at now is a big jump and would match the increases that Intel got with the current process size for the Pentium 4. Also, IBM states that the companies upcoming 970 powered blade server will go up to 2.5 GHz. So, you are effectively saying that Apple will get a 970 that is 700 MHz beyond the maximum that IBM will use in that blade computer. If a jump up in volts is what is needed to get to 3.2 GHz, then that would be a 27% increase in frequency over the topend IBM blade server speed. That would be highly unlikely to occur also.

[jettredmont]

Discussion thread recap ('cause it spans about five pages):

Analog Kid said:

Look at how quickly amateur coders are putting out Cocoa apps. If you have more cycles to burn, you can focus less on optimizing and more on putting apps together quickly with more reusable code.


to which Rincewind42 replied:


Sure, it's great to be able to put an app together without writing more than a handful of code - but if you can do it so can thousands of others and your app really isn't all that distinctive in that sense then is it? A distinctive app requires a programmer to take full advantage of all of the tools available to him or her - including optimizing.


to which I replied (specifically at the first sentence above, not the second; I agree that optimizing is one of many tools, but not that a lack of optimization means an undistinctive app):


I disagree. A distinctive app requires more than programming prowess. It requires intelligent design, from the UI to the inner workings.


And now, back to the game ...

Originally posted by Rincewind42
And to optimize you must know that you are using the best algorithm. If you aren't already using the best algorithm, then you are wasting your time optimizing suboptimal code.


In my field, there is never "the best algorithm". There are a hundred different theoretical approaches (and more proposed monthly), and some of them may in some cases be far better than others (and in other cases far worse). Most of these approaches have never been tried in code yet.

So, I admit, this skews my thinking a bit. Optimizing a particular implementation of a particular algorithm is good, but more often than not getting the job done fast is more a matter of being flexible and allowing experimentation than honing an implementation to as sharp an edge as possible.

In other words, back to my original post: building an app out of mostly-reused code does not mean that "anyone" else could do it just as well. In some fields, yes it obviously does (but then again, "anyone" can optimize their code too). But in fields where fundamental innovation is still possible, where the race amongst competitors is one of approach and elegance of implementation, reusing code, even though it is not 100% optimal for your approach necessarily, is the difference between the leaders and the followers.

I mean, yes, I could be getting a good 20-30% gain by using C constructs (structs and arrays instead of classes and vectors) throughout. In fact, my project has a significant portion which remains straight C because we've never bothered to go in and change it. Is the remaining C code a performance advantage? No, far from it. It is the slowest section of our application, because it is poorly understood and thus never innovated. The rest of the application has a record of 100% performance gains year-over-year. That is not from low-level optimizations; that is from devising and implementing (quickly) new approaches to the problem at hand. While this section of code started off as the fastest bit in the project, the C code rarely gets faster, and so, year by year, becomes more and more the bottleneck of the system. I know the C vs C++ speed advantage because I'm finally rewriting this chunk and, yes, a direct rewrite with no approach change does slow that code down by a good 25%. Said rewrite, however, will allow this bit to improve in the same way the rest of the app has improved over the years (and in fact we have a few ideas to try here that look like they'll make the new code just as fast as the old code was ... ideas we'd never have dreamed of implementing on the old C code because it just plain would have taken too much time).

For the record, much of this same conversation took place on one of Apple development lists last week. That discusion concluded that optimization is a necessary step in development, not just so that things work well on older hardware, but also so that they work well on newer hardware.

Of course one has to make sure your code works on the lowest target machine. That doesn't mean you have to develop on a 400MHz G3 clunker, but it does mean you should sanity-check and monitor ongoing performance on such a machine. And, yes, optimizations have to take place (to a certain extent) to achieve this. As I said before, however, performance is not only affected by assembly-level optimizations (by which I mean looking at the generated assembly and refining your C code to generate five instructions instead of eight). It is often (in my experience, at least) more affected by higher-level optimizations. AND such high-level optimization are more likely to work across multiple platforms than any assembly-level optimization!

My application runs well on an old G3 (reasonably so; it scales pretty linearly with processor speed; it runs faster than the competition on all hardware) as well as half-decade-old P3s because we have used off-the-shelf modules and high-level ingenuity to quickly implement fast algorithms, not because we have unrolled loops and used pre-increments instead of post-increments and put loop variables on the stack instead of from the heap, etc.

I'm not against low-level optimizing. I just don't agree that it is the key differentiating factor between applications. It can differentiate, but the absolute lack of assembly-level optimizations does not mean your application is undifferentiated.

At some point, you have to review and clean up code, rethink your low-level approaches and, and run it through Sampler to get an idea of exactly where you are spending time. That doesn't mean you need to be writing your own implementation of vectors and hashtable maps. Using "stock" code does not mean you get poor performance.

[jettredmont]

Originally posted by eric_n_dfw
very good points - maybe the PB G5 is farther off than I initially predicted. As far as the actual wattage of the G5's, can anyone point me to an official (Apple or IBM) page that lists them? All I can find is some article that states the 2Ghz G5 is a whopping 97 Watts.

Numbers are hard to come by, but EE Times is fairly respectable (ie, I don't imagine they made up the 97 Watts thing ...)

http://www.eetimes.com/sys/news/OEG20030623S0092

[bcsimac]

I hope that Apple and IBM can continue to work out for the both of them. I think that IBM is a better innovator than motorola. I also think IBM has much more resources to work with than Motorola has. The problem is making sure that IBM continues to think it is profitting from helping Apple with its processor needs. I think right now...it is all good. I just wonder how long it will stay that way. Apple and IBM have parted ways before at least twice. I would hate to see that again. I just really hope that this time they will stay in it together and not part ways because IBM is better than Motorola if you ask me.

[rickag]

Originally posted by jettredmont
Numbers are hard to come by, but EE Times is fairly respectable (ie, I don't imagine they made up the 97 Watts thing ...)

http://www.eetimes.com/sys/news/OEG20030623S0092

No they were confused. IBM has published at least 2 documents that contradict this, by a very large margin. Go to IBM's website, look up the 970 documentation and see for yourself.:D

[Rincewind42]

Originally posted by rickag
No they were confused. IBM has published at least 2 documents that contradict this, by a very large margin. Go to IBM's website, look up the 970 documentation and see for yourself.:D

Could you please post a link to these documents? I think that since everyone saw the heatsinks they assumed that IBM completely blew their power specs =p.

Btw: if your looking at the article & the product presentation, that was pre-release specs, so they don't really count anymore :(

[Frohickey]

Manufacturers blow power specs all the time.

I remember a chip that, when it came out was TWICE the power on the datasheet. After that, you revise the datasheet...if you remember. Engineers are usually not that good with documentation, and documentation is one of the first things to go when you get busy.