Upcoming Intel Details Leak

[Phinius]

Re: Re: Apple will never catch up in GHz, worse yet....

Originally posted by Abstract


The only thing I disagree with is that Apple will be without a G5 in their PB line until January at the earliest. Sorry, but they could just fit a 1.2GHz G5 in there and appease the masses.

If Apple fit a 1.2 GHz G5 in the PB line, then there would not be another update for about eight months. There simply is no more speed to give for the G5 in a notebook computer at the process size that it is at now.

It looks like Apple will wait until a few months before the next process shrink for the G5 to introduce the G5 into PowerBooks.

Nobody is asking Apple to stick in a dual proc 2.0GHz, just a 1.2GHz G5 that can compete with the Centrino.

PowerBooks with a G4 running at 1.5-1.6 GHz should be quite competitive to the Centrino line. Look for Apple to introduce PowerBooks this month that are a major step up in speed with the smaller process size G4s.

I'd also expect Apple to put as much as a 1.8 GHz G4 in the iMac line starting this month also, although 1.8 GHz maybe a bit optimistic at this point in time. At least the 1.42 GHz of the Power Macs though.

 

[acj]

3.4GHz at 130nm

Originally posted by Macrumors

Prescott is to start at 3.4GHz and expected to cost $637 in volume quantities. New Celeron and Pentium M (laptop) processors are also expected later this year. The Pentium M will feature a 90-nm manufacturing process.

Just for those who missed it. The Pentium M will be made on the 90nm process. The first Prescotts will be on 130nm. This gives lots of room for improvement with the move to 90nm.

 

[Abstract]

Re: Re: Re: Apple will never catch up in GHz, worse yet....

Originally posted by Phinius
If Apple fit a 1.2 GHz G5 in the PB line, then there would not be another update for about eight months. There simply is no more speed to give for the G5 in a notebook computer at the process size that it is at now.

It looks like Apple will wait until a few months before the next process shrink for the G5 to introduce the G5 into PowerBooks.

Yes, updates would still be possible. If they were to introduce a 1.2GHz G5 into the PB, it would tie over Apple users for a while, say 6-8 months. In the meanwhile, IBM will introduce a 970 using the 90nm process. It will be introduced within the year. Apple can then update the PB's with a G5 using at the 90nm process in 6-8 months.

Maybe I'm mistaken, but didn't Steve say that chips produced at 90nm will be available within the year, and that the 970 will scale to 3.0GHz in 12 months at the 90nm process? I thought the 90nm process was going to be introduced in the 970's within the year (when the 970's were at 2.5-2.6GHz) and carried over to the 980's in 1 year or 1 year and 6 months from now when they're introduced. I only said to introduce the 1.2GHz G5's into the PB's under this assumption. At 90nm, fitting a 1.8GHz G5 into a notebook shouldn't be much of a problem, I'd reckon. Let it happen in 6-8 months.

Of course, I could be completely wrong and talking out of my ass.

 

[SiliconAddict]

Apple and IBM are going to have to seriously stay on their toes to maintain that lead they have over AMD and Intel. One thing worries me. Most of the info on Intel's chips are on the desktop side where Apple already has a strong advantage. Everything I’ve read says that Intel’s mobile Pentium M chip is going to 90nm with a 2MB cache and a 400Mhz system bus in their laptop line this fall. As I had commented in a previous thread I have yet to see ANY benchmarks of a 17” PB going head to head with a Pentium M. Even if the Mhz myth is true the fine tuning Intel is doing (Cache, system bus, etc.) could widen the gape in performance unless Apple\IBM\ Motorola pulls off a Houdini. The laptop line needs to be shored up. Its really that simple.

Honestly guys. Intel doesn’t even really care about the G5. As cool as the title “fastest computer” is it doesn’t realty affect their sales all that much. Its the whole damn WINTEL thing. If a G5 could be put into a PC architecture then you would have them worried. Its AMD and to a lesser extent Transmeta who they are focusing on. These pot shots AMD and Intel continually throw at each other to be the “speed champ” has, until recently, left Apple\IBM\ Motorola eating vapors. I will be highly surprised if by fall of 2004 Apple and IBM still holds the speed crown.
Don’t get me wrong guys. I’m rooting for Apple and IBM. I’d be more then happy to eat crow fall of next year. I’m just saying they have a tough fight ahead.

 

[Phinius]

Originally posted by acj


And currently available top end consumer Macs use more power than currently available top end consumer PC's.

I sincerely doubt that.

Two G4's generate more heat than one P4. Unfair comparison? Well the performance is comparable.

I would put the power use of two topend G4s at about the same as a topend Pentium 4 and no the Pentium 4 will not be twice as fast as a G4 starting this month. In fact a Pentium 4 won't even have close to twice the performance of a G4 starting this month.

The G4 is highly underrated as a processor. A comparison of a topend Pentium 4 to a dual-processor G4 PowerMac does not have two G4 processors running full throttle. The additional processor never gives the PowerMac more than a 50% boost in speed over a single G4 running at the same frequency. Comparing a single 1.42 GHz G4 to a 3.2 GHz Pentium 4 processor will not result in the Pentium 4 doubling the performance of the G4.

 

[Phinius]

Re: Re: Re: Re: Apple will never catch up in GHz, worse yet....

Originally posted by Abstract
Yes, updates would still be possible. If they were to introduce a 1.2GHz G5 into the PB, it would tie over Apple users for a while, say 6-8 months.

Apple's Jon Rubenstein has already stated that the G5 will not go into the PowerBooks for quite some time. That's probably due to the new G4 having superior power use/performance compared to the G5 on the .13-micron process. When the G5 moves to the smaller .090-micron process it will be a whole different story when comparing a G5 to a G4 for power use/performance in a notebook computer. Right now though the G4 should outclass the G5 for power use/performance in a notebook computer.

Maybe I'm mistaken, but didn't Steve say that chips produced at 90nm will be available within the year, and that the 970 will scale to 3.0GHz in 12 months at the 90nm process?

Steve said nothing about future process shrinks for the G5. He simply stated that the G5 will be at 3 GHz within a year.

I sincerely doubt that IBM will move the current 970 processor to the smaller .090-micron size. I'd expect the 980, which should be based on the upcoming POWER5, to be the next G5 on the smaller process size. I certainly could be wrong, but if the 970 is moved to the smaller process first, then the 980 will trail the introduction of the POWER5 by quite a few months. Afterall, why introduce the 980 a few months after the 970 was moved to a smaller process size? It wouldn't be very economical to do that.

 

[Sol]

No big deal

For all the hype about the new Pentium, it fails to impress the way the G5 PowerMacs do.

I heard that Intel is minimizing the over-clocking of its CPUs and that would certainly be in contrast to the PowerMac's design which seems almost tailor made for over-clocking; louder, like the last generation of G4 PowerMacs. Each G5 has its own fan and that could mean that the CPUs will be over-clocked when required. Apple already does this with the G4 PowerMacs which are rated higher than Motorola's fastest 1.2MHz G4s. When this new Pentium is released Apple could update its PowerMacs with "faster" processors.

More cache does not impress me as much as no cache and the G5 has none. New Pentium? Big deal. It seems to me that Intel has hit the x86 wall and the only thing that can help it is an even smaller fabrication process. 0.0 nanometers maybe?

 

[acj]

Originally posted by Phinius
I would put the power use of two topend G4s at about the same as a topend Pentium 4 and no the Pentium 4 will not be twice as fast as a G4 starting this month. In fact a Pentium 4 won't even have close to twice the performance of a G4 starting this month.

Your estimates of power may be closer to the truth, however what do you mean, "starting this month?"

Originally posted by Phinius
A comparison of a topend Pentium 4 to a dual-processor G4 PowerMac does not have two G4 processors running full throttle. The additional processor never gives the PowerMac more than a 50% boost in speed over a single G4 running at the same frequency.

Exactly. One fast CPU is better than two slow ones. Two fast CPU's are better than one fast CPU (Tada! G5 DP 2.0!)

Originally posted by Phinius
Comparing a single 1.42 GHz G4 to a 3.2 GHz Pentium 4 processor will not result in the Pentium 4 doubling the performance of the G4.

Not exactly. The 3.2 P4 beats even the Dual G4 1.42 in many tests, including fair photoshop tests. Besides, why would you compare a non-existent system to a top of the line system?

 

[VIREBEL661]

I feel there's a place in the world for Intel AND PowerPC chips... I prefer PowerPC and RISC technology, as well as my favorite operating system, however, if I didn't have Mac OS, I would run Linux on a PowerPC, then, without that option, Linux on Intel hardware... Like most of us on these boards, I think it's pretty safe to say that windoze is the most loathesome part of the PC world... If Intel builds ever faster procs, then so be it (they're not RISC processors, so sort of a different technology, this came up a lot when Mac went PowerPC)... It doesn't change what my favorite hardware is, and actually will promote faster growth in processor speed from IBM, who, no matter what, isn't going away (neither's Apple, though some ignorant folks like to think otherwise, I mean, c'mon, they INVENTED the personal computer, and perfected the GUI to be ripped by M$, jeeeeeesh, even Bill Gates admits that)... I think the next couple of years are going to be exciting for we Mac users! Stay tuned....

 

[Rower_CPU]

No political discussion allowed in the news thread.

 

[sinclairZX81]

Re: Apple will never catch up in GHz, worse yet....

Originally posted by websterphreaky
[BJobs is already learnig to lie in public statements, apparently from the supreme liar, Al Gore. [/B]

I'll start worrying the day Shrub says that Apple has Weapons of Mass Gigahertz up its sleeves... 'cause then you KNOW it'll turn out not to be true.

You spelled 'golden' incorrectly.

 

[Docrjm]

Originally posted by mgargan1
oh, and they're not going to name it pentium 5, cause pentium already means 5. And it would be a little redundant calling a processor 5 5. The pentium name came about when the courts would not allow intel to copyright the numbers 386...486... etc. So they have to do some thinking. I believe they'll keep with the pentium 4 moniker; however, it wouldn't be a smart idea, cause the name's been out for almost 3 years.

No dafter than pentium 4, using your logic that is 5 4.

Don't matter what it is OS X still smokes xp.

 

[Scottgfx]

Remember what Jobs said?

I was thinking back and did a google search to find this.

(From MacNN: July 18, 2002)
As the megahertz gap widens between Intel-based PCs and the latest PowerPC Macs, some analysts are urging Apple to reconsider its microprocessor strategy. When asked if Apple would seek an alternative to Motorola's PowerPC, namely chips from Intel, Apple CEO Steve Jobs said the company first had to finish the transition to Mac OS X: "Then we'll have options, and we like to have options," he said.

I think Jobs just showed us the options last month. The words "we like to have options" have stuck in my head. In the meantime I had to deal with a lot of people asking me about John Devorak's (sp?) article on Apple moving to Intel processors. For some reason I felt pretty confident in Apple using the PPC970. I'm also still waiting for the next Amiga.

 

[fpnc]

Re: Re: Re: Apple will never catch up in GHz, worse yet....

Originally posted by Phinius
PowerBooks with a G4 running at 1.5-1.6 GHz should be quite competitive to the Centrino line. Look for Apple to introduce PowerBooks this month that are a major step up in speed with the smaller process size G4s.

I'd also expect Apple to put as much as a 1.8 GHz G4 in the iMac line starting this month also, although 1.8 GHz maybe a bit optimistic at this point in time. At least the 1.42 GHz of the Power Macs though.

I also think it is possible that we'll see an improved and speed-bumped G4 in the PowerBook fairly soon. However, a jump to 1.5-1.6 GHz in a G4 PowerBook seems way too optimistic and a 1.8 GHz G4 in an iMac will probably never happen (at that level of performance it would probably make more sense to use a G5). I'd say that 1.2 to 1.4 GHz is all that we should expect in a new G4 PowerBook and/or iMac. In fact, that may be where the G4 and Apple finally go their separate ways (i.e. the end of the G4-based Macintosh).

 

[AhmedFaisal]

What's the thermal design power going to be? 100 Watts?

The current P4 has a thermal design power of solid 82 Watts and the maximum allowed temperature is 72°C. Boys, that's a server CPU for my taste. You need expensive watercooling or a supersonic fan to keep that beast cool. I doubt its going to get any better with the new core. The P4 is a failed design and Intel has already admitted it by designing the Pentium M which has a much better GHz/Processing Power ratio. I for one am laughing my ass off at all those "real GHz" morons with their noisy boxes and laptops with exhausted batteries after 30 minutes. There is at the moment only 2 decent chip design out in the x86 world, the Pentium M and the VIA C3, everything else (including AMD) is overprized cooking plate competition.
Cheers,

Ahmed

 

[Analog Kid]

Originally posted by shawnjackson
oh and anyone who keeps saying that intel chips run hot should actually use one. I have two dp 1.42, an athlon xp 2100+ and a 1700+ and a 3.0 p4 c...

the g4 and the athlon are probably tied for "hotness" at around 55 deg C at max load (using a thermaltake heat probe at cpu core.

the p4 runs load at about 38 with a nice quiet zalman fan. thats why its my home theater system.

Tell me you compensated for the heat sinks in this test...

I have no difficulty imagining the temperture of a G4 and Athlon are the same with their heat disipation hardware in place. That's kinda the point...

 

[eric67]

Price, frequency and Intel road map

Originally posted by scem0
It is true that the true battle will be over software, but the more glorified battle (if less important) will be the speed battle.

It is also a stupid battle.

They are both plenty fast.

If apple needs to work on one thing it would be price. It is pretty hard to justify spending $3,000+ on any computer (for most people).

scem0

This is true, and this probably the main reason why Apple share is still around 3-4%. The G5 will only push mac user to exchange their getting-old G4 to brand-new G5. But the real goal for Apple is to concentrate on normal consumer, who do not know so much about computer and just want a computer to use it (which now days means, internet, finance/budget, e-shopping, games- the main problem for Apple, Apple should push or invest money to help developers to release new games for both platform at the same time- education) and now on what are the main driving forces??? Price ad GHz myth!!!!!!!!!!
Price : well no need to explain, yes I know that a iMac is better equipped than a entry-level PC, but the latter is about 2 times cheaper.... those interested by a low end PC do not care not having FireWire,.....

GHz myth : yes I agree that this the main strength for Intel, increasing the frequency makes people thinking that it goes faster, well if you look at the Prescott P4, it will top at 3.4GHz, indeed if you increase frequency you will not see any gain for performance. Just look at the gain obtained between a P4 3.0 and 3.2 GHz, it is nothing!!!!!!!!!111
So then will come Tejas, well to be able to keep going with frequency, then Intel has planed 26 steps pipeline for Tejas , the follow up, Nehalem, will have 38 steps pipeline. in other world to keep with performance, Intel will have to increase frequency, and the only way to do it is to have the smallest process coupled to high frequency. .... but the real gain on performance will of course be lower than the GHz increase. Why do you think Intel introduce hyper threading??? because they know that they are getting close to the top frequency for the P4 Northwood!!!!!!! (3.4 GHz)
so they are looking to other way to increase performance, than GHz increase.....
NOW, all this discussion on GHz myth, pipelines, frequency/performance, the standard consumer does not give a ****, what is looking at is his wallet, and if his wallet says I can get a computer, PC-type, for less money that the iMac, and you can do more or less the same thing (not speaking video or imaging processing of course), then why getting a Mac???

Here is the point that Apple has to solve, and not only rely on their pro consumer, I know so many people who knows Mac, would love to get one, but can not afford one ......
I have been among then for a quite long time.

 

[AhmedFaisal]

Originally posted by shawnjackson

oh and anyone who keeps saying that intel chips run hot should actually use one. I have two dp 1.42, an athlon xp 2100+ and a 1700+ and a 3.0 p4 c...

the g4 and the athlon are probably tied for "hotness" at around 55 deg C at max load (using a thermaltake heat probe at cpu core.

the p4 runs load at about 38 with a nice quiet zalman fan. thats why its my home theater system.

First thing I did on my dual 1.42 was switch the fans and especially the CPU cooling system to that of a 3rd Party, now its quiet like my baby son when he sleeps. At any rate, P4 is one lousy selection for that home entertainment system. Since Apple is still denying us a pizzabox headless model I look what I could find in the PC world and this is what I use: www.hush-technologies.com. Its a neatly designed machine with a VIA Epia M 10000 ITX board, a 120MB sound insulated seagate barracuda and a slimline combo drive. All passive cooled so no noisy fans involved and so small it fits perfectly to my stereo and looks absolutely cool too. Its plenty fast for DVD & DivX and AAC audio music files. Only thing I added extra was a M-Audio Audiophile since I prefer an audio card with CINCH output and the sound of the audiophile is simply pristine.
Cheers,

Ahmed

 

[soggywulf]

Re: No big deal

Originally posted by Sol
It seems to me that Intel has hit the x86 wall...

Yeah right. We've been hoping for that since the 601. Hasn't happened yet. CISC and RISC are just a bit of microcode apart. Intel has plans, and they'll stay in the game or ahead of it.

Besides, why would anyone actually want Intel to hit a wall? If that happens, then IBM lays off the amphetamines and our Macs won't get faster and cheaper. That would kind of suck wouldn't it?

 

[tazznb]

Never let your guard down...

Originally posted by rjwill246
You aren't thinking that a 3.4 Gig Pentium 4 is faster than say, a 2.5 Gig G5 are you? You wouldn't be right.
And to the other correspondent: yes it is possible to fry eggs on maxed out P4s... stunning, over easy! and you can have the chips with it too, though fish is better.

Actually these new 3.2 p4's can go as high as 100W (cooks steaks well done).

I've heard that the P5 (Pentium 5) will be very fast (the next chips 3.4Ghz coming from the great whore Intel will indeed be P5's).

For your reading pleasure:
http://lowendpc.com/rumormill/2002/0709.html

 

[soggywulf]

Originally posted by scem0
It is also a stupid battle.

They are both plenty fast.

No, they ain't.

 

[ewinemiller]

Originally posted by SiliconAddict
As I had commented in a previous thread I have yet to see ANY benchmarks of a 17” PB going head to head with a Pentium M. Even if the Mhz myth is true the fine tuning Intel is doing (Cache, system bus, etc.) could widen the gape in performance unless Apple\IBM\ Motorola pulls off a Houdini. The laptop line needs to be shored up. Its really that simple.

Actually Barefeats has a small blurb of a pb 17 vs. a 1.3Ghz Centrino laptop. The end result is that the Centrino ran CineBench 100% faster, Photoshop 20% faster, quake at 640x480 103% faster. The PB gained the upperhand with quake at 1024x768 by 84% because of more video memory.

Here's the link, scroll down to 6/28/03

http://www.barefeats.com/#quick

 

[ewinemiller]

Originally posted by scem0
It is true that the true battle will be over software, but the more glorified battle (if less important) will be the speed battle.

It is also a stupid battle.

They are both plenty fast.

If apple needs to work on one thing it would be price. It is pretty hard to justify spending $3,000+ on any computer (for most people).

scem0

For many folks they are not plenty fast. For my mom, yes, she just recent upgraded from her Pentium 133 laptop to a bright shiny new Pentium II 366 and only because the 133 couldn't install enough memory to run turbo tax. BTW, for all the folks that keep saying a PC is useless after 3 years, the 133 is about 8 years old and the 366 is 4. Her husband is running my old dual 266, overclocked to a screaming 300, that is nearly 6 years old.

Anyways back to the original point, for me, the need for speed is always growing. When I had my P166 I used to let my 3D renderings run over the weekend, that was with ray tracing. As CPUs got faster, I didn't have to do that anymore for stills, but now I'm rendering with photon mapping and on some complicated scenes I'm back to letting my renderings run over the weekend. I need faster machines, speed is very important to me.

I am very curious how well the G5 and the new prescotts will do as I start to think about a new machine, but I'm not going to base my decision off any benchmark that isn't what I do. I've seen both platforms do very good at one thing and fall down on the next. I've even seen significant variation base on the particulars of different test files for the same application. I'll hopefully talk the first sucker, er, happy new owner of each to run a few files off the content CD for my particular package. With those times I can compare to what I get now and then make a decision.

 

[wizard]

Hi Silicon Addict;

I'm somewhat in agreement with what you say here Apple will have a very hard time keeping up with Intel performance wise. In fact I'm not willing to say that they will or now have a performance lead with the G5. The G5's performance as deomonstrated by Apple is a mixed bag, many of the so called bake offs and benchmarks seemed to be picked to demonstrate band width, not CPU performance. Not every operation that the average user does on a computer is bandwidth limited.

I also agree that Apples laptop line is hurting big time. They need a substantial increase in in G4 performance on these machines, much of that requires a bandwidth increase that apprently Motorola will not release anytime soon. The current hardware implementation for the G5 though has an huge amount of bandwidth but that leads to heavy power disapation outside the main CPU, which is not good for battery hold up time. It will be interesting to see if Apple has implemented a differrent chip set for the G5 designed for low power usage in a laptop or IMac. The other problem is that the G5 would have to be implemented at a rather high clock rate in a laptop to really compete with a G4. The only advantage that a G5 running at 1.2 GHz in a laptope would be the addressing capabilities, in most other ways newer 1.4 GHz G4's would beat it.

Laptops demonstrate clearly that if Apple has compelling technology they can be very successful selling their hardware. The problem is they really need to stay ahead of the Intel/AMD crowd to maintain that advantage. The Centrino package is something to be concerned about.

While I don't believe in the MHz myth Apples hardware has lagged greatly in performance improvements over the last couple of years. It is important to realize that speed isn't always the ultimate judge of performance of a laptop, on the other hand it wouldn't hurt Apple to have one performance laptop in their lineup.

Dave

Originally posted by SiliconAddict
Apple and IBM are going to have to seriously stay on their toes to maintain that lead they have over AMD and Intel. One thing worries me. Most of the info on Intel's chips are on the desktop side where Apple already has a strong advantage. Everything I’ve read says that Intel’s mobile Pentium M chip is going to 90nm with a 2MB cache and a 400Mhz system bus in their laptop line this fall. As I had commented in a previous thread I have yet to see ANY benchmarks of a 17” PB going head to head with a Pentium M. Even if the Mhz myth is true the fine tuning Intel is doing (Cache, system bus, etc.) could widen the gape in performance unless Apple\IBM\ Motorola pulls off a Houdini. The laptop line needs to be shored up. Its really that simple.

Honestly guys. Intel doesn’t even really care about the G5. As cool as the title “fastest computer” is it doesn’t realty affect their sales all that much. Its the whole damn WINTEL thing. If a G5 could be put into a PC architecture then you would have them worried. Its AMD and to a lesser extent Transmeta who they are focusing on. These pot shots AMD and Intel continually throw at each other to be the “speed champ” has, until recently, left Apple\IBM\ Motorola eating vapors. I will be highly surprised if by fall of 2004 Apple and IBM still holds the speed crown.
Don’t get me wrong guys. I’m rooting for Apple and IBM. I’d be more then happy to eat crow fall of next year. I’m just saying they have a tough fight ahead.

 

[eric67]

Originally posted by ewinemiller

I am very curious how well the G5 and the new prescotts will do as I start to think about a new machine, but I'm not going to base my decision off any benchmark that isn't what I do. I've seen both platforms do very good at one thing and fall down on the next. I've even seen significant variation base on the particulars of different test files for the same application. I'll hopefully talk the first sucker, er, happy new owner of each to run a few files off the content CD for my particular package. With those times I can compare to what I get now and then make a decision.

well us.macbidouille.com reports a test which has been performed by the NASA regarding the G5 vs P4
actually the G5 with no optimized code did really well.
article : http://www.hardmac.com/niouzcontenu.php?date=2003-07-04
direct link : http://members.cox.net/craig.hunter/g5/