Update: Apple and the PowerPC 970... Next Year?

[MacRumors]

This ZDNet article (http://news.zdnet.co.uk/story/0,,t269-s2131244,00.html?rtag=zdnetukhompage) recaps IBM's press release on the upcoming PowerPC 970 (covered here (http://www.macrumors.com/pages/2003/02/20030227065458.shtml)).

The 970 processor is widely expected to make its way into Macs... and ZDNet's sources agree, but perhaps not until next year:

Sources familiar with IBM and Apple's plans said that Apple will be a customer for the PowerPC 970 next year, but Apple has declined to comment.

MacRumors has heard that upper-level Apple Executives are worried about the number of PowerPC 970 rumors circulating -- with concerns about current PowerMac sales in the light of these rumors.

Per MacRumors sources, there have been conflicting rumors about the timing of the 970... but due to the unpredictable nature of product development, it's possible that the release date may simply not yet be decided.<P>

Update: ZDNet has the same quote (http://www.zdnetindia.com/biztech/enterprise/computing/stories/68498.html) regarding "next year" in an article from October 2002.

[Performfreak]

I think that apple should just come around and confirm whether the 970 is going to be used or not... If their sales are already being hurt due to rumors and assumptions, why don't they just confirm or deny what we are all so curious about.

I think then that those who need a good computer now will buy one, as the new powermacs have life in them left due to the dual config and everything, and then those who would want to wait for the 970 (most, probably) will be satisfied. I think it's about time to put the G4 powermac to rest, we've had it for far too long, and apple should just announce the 970 despite the possible slowing sales of the G4s.

The sales have gradually slowed whether there has been the anticipation of a new product or not. The the G4 powermac is becoming a less competitive product, simple as that. Don't get me wrong, I own a MDD and love it, but it's time to put the G4 in the consumer level permanently and the 970 (or other possible replacement) in the Pro level.

[agreenster]

Well there you go. Hopefully this will help spur some PowerMac sales.

[Freg3000]

I'd rather believe the Page 2 rumor that says it'll be available in July. :(

[yzedf]

MacRumors has heard that upper-level Apple Executives are worried about the number of PowerPC 970 rumors circulating -- with concerns about current PowerMac sales in the light of these rumors.

PowerMac sales suck because the PowerMac sucks. That is why the rumors are so strong. We (Apple's target market) realize the need for a real update, not some speed-bump to the anemic G4.

[Hemingray]

Yup, I'll continue to stand my ground on this... we'll see PPC970 no sooner than Jan. 2004. 2003 is the year of the "laptop", after all...

[testnull]

ZDNet strikes again!

Their story from October which recapped the MPF reports said:

"Sources familiar with IBM and Apple's plans said that Apple will be a customer for the PowerPC 970 next year."

Their story from this month which recaps the Blade server rumors says...

"Sources familiar with IBM and Apple's plans said that Apple will be a customer for the PowerPC 970 next year."

This is cut-and-paste gerbilism, er, journalism. Move along, folks.

[Kwyjibo]

hmm if apple confirmed that they would use the 970 and on x date even less people would buy the curent pmac or only the ew who need it. In fact I think that if apple doesn't have plan to use it this year they shoudl make ap ress release to that effect fro those consumers that are one the edge.

[MacsRgr8]

To be honest, I'm glad I bought my new Dual 1.25 GHz two weeks ago....
The waiting for the PPC 970 seems endless. Last X-mas I promised myself to wait until the 970 came out. But MWSF got me a bit nervous. I didn't want to wait another whole year before replacing my 3 year old G4 400.
What bothers me furthermore, is that the first Mac ever to see (hopefully, IT IS STILL NOT CONFIRMED) the 970 could be a new Xsere, or WorkStation PowerMac (>= $ 4000)... you know what I mean.
Still the sooner the better!

[arn]

It's true...

"Sources familiar with IBM and Apple's plans said that Apple will be a customer for the PowerPC 970 next year. Apple did not return calls for comment. "

http://www.zdnetindia.com/biztech/enterprise/computing/stories/68498.html

From October 2002.

Oh well.... sorry for the false alarm...

arn

[Mr. Anderson]

ah, please hope it isn't so. And if its all a ploy to boost sales of the current crop of machines, fine. It won't make me go out and buy a new one, I'm waiting for the next generation, period.

Regardless, it'll mean the wait will probably be longer. :(


D

[macktheknife]

Originally posted by Performfreak
I think that apple should just come around and confirm whether the 970 is going to be used or not... If their sales are already being hurt due to rumors and assumptions, why don't they just confirm or deny what we are all so curious about.

I think then that those who need a good computer now will buy one, as the new powermacs have life in them left due to the dual config and everything, and then those who would want to wait for the 970 (most, probably) will be satisfied. I think it's about time to put the G4 powermac to rest, we've had it for far too long, and apple should just announce the 970 despite the possible slowing sales of the G4s.

The sales have gradually slowed whether there has been the anticipation of a new product or not. The the G4 powermac is becoming a less competitive product, simple as that. Don't get me wrong, I own a MDD and love it, but it's time to put the G4 in the consumer level permanently and the 970 (or other possible replacement) in the Pro level.

I couldn't agree more. PowerMac sales have been fallen by more than half since 2000. While the poor economy is partly to blame, the chasm in Mac-PC performance is discouraging potential customers to shell out big bucks for a slower machine. At this point, Apple should just let everyone know what they plan to do: if they're coming out with the 970 next year, fine. Those who really want a PowerMac now will buy it, while those who want to wait will get an idea of how long they need to wait.

[dricci]

I have no doubt in my mind we'll see it this year, Apple can't afford to wait any longer - and they know that. I'm sure that they have Mac OS X 64-biterized and ready to run on these things, or are at least at the final development stages of porting it. Apple's probably ready, but they can't do much until IBM has the chip finished and ready to be mass produced.

[chubakka]

Now everyone is gonna freak. hehe...

I'm thinking 2nd half of this year.

Hoping for ship dates in August.

[deputy_doofy]

Yes, well I like to add to Apple's uncomfortability by sending them feedback messages telling them how much money I have saved up just waiting for a new 970-based machine. I can FINALLY afford a top of the line PowerMac (though I have no business need for one) but they're just not getting my money until I like the hardware. I love my iBook and it will carry me through the months until a [REAL] new system appears.

[Dont Hurt Me]

If apple thinks forcing g4s on more people is going to increase sales they are dead wrong. Even now it takes two g4's just to get on the same page as 1 p4 and they still come up short! The longer they wait the longer they screw themselves. Then again maybe they have bought a zillion g4s from slow butt motorola and now have to do something with them. Maybe they should hire the macrumors crowd to plan their cpu strategy?

[gopher]

While 64 bit processors will be faster than the current crop of G4s, people are deceiving themselves if they think the current crop of G4s is always slower than Pentium IVs.

http://forgetcomputers.com/~jdroz/09.html

The above link shows that in the millions of theoretical floating point calculations a second (MTOPS) the G4 has the Pentium beat 38:8. Even the iMac/eMac are faster than the fastest Pentium or AMD. So why are we seeing tests from Adobe Premier beating the G4? Software optimization. Premier's Altivec coding is not as good as it could be.

http://www.macdevcenter.com/pub/a/mac/2002/04/05/altivec.html

shows how to code it properly. When it comes to RC5 and Genentech Blast, the G4s have the Pentium beat 5:1.

So it all comes down to people not developing altivec code into their software. Oh and if Mac OS X appears to be slow for you, you may not have been doing enough to keep it optimized:

http://www.macmaps.com/Macosxspeed.html explains how to do that.

If you have been waiting for a fast enough G4 for a long time, you'll be surprised how fast the current crop of G4s are. And by buying it now, you are in fact saving yourself time in the gain of performance you get from a new machine today. How can you save yourself time waiting for a new machine? All you are doing is posting here your complaints that the G4 is not fast enough. If you just got a G4 and worked with it, you'd get more work done quicker! Performance isn't all processor speed. Performance is not procrastinating and getting what you need done today!

[Dont Hurt Me]

Great gopher,5% of the software makers know how to code for the g4, bottom line is the other 95% are getting their butt kicked by p4. when almost every app is getting blasted by a cheaper machine its time for apple to do something. Its time for the 970 now not next year sometime, almost sounds like motorola and their upgrade path. Its gotten so bad apple cant even compare that one gauzeen blurr they use to use. Apple still has the best software,best style, but slowest hardware. Go look at any game or photo app and the truth is the g4 is getting smoked and thats why apple resorted to two of these cpus to compare with one intel.DROP I DONT CARE MOTOROLA AND DO IT NOW!:o

[gopher]

It doesn't have to be that way. If people started writing their favorite software developer for better Altivec coding with:

http://www.macdevcenter.com/pub/a/mac/2002/04/05/altivec.html

and showed them it really is worth it considering how good the floating point performance is:

http://forgetcomputers.com/~jdroz/09.html

People would buy their software more often. Now that iMac, eMac, Powerbook, and PowerMac are all G4 machines, the only thing that's stopping software developers from developing better Altivec coded software is the demand from users. If you haven't demanded it, you should. You are more likely to get an optimized piece of software than a PowerPC 970. By the way, since the PowerPC 970 derives some of its strength from Altivec coding, once again, we are going to see that unless developers develop for Altivec, the true strengths of the PowerPC won't be realized. The sooner you start writing your favorite developer to optimize their code, the better for all of us. It starts with user demand. With none, you don't find better software.

[3G4N]

Hannibal hinted that his second installment
of his 970 article over at ArsTechnica (http://www.arstechnica.com),
is due out early next week!

[bwawn]

Originally posted by Dont Hurt Me
Maybe they should hire the macrumors crowd to plan their cpu strategy?

I hope they wouldn't hire the Macrumors crowd for anything else. They'd go bankrupt with supplies of millions of video iPods and other difficult-to-market products...

[Sol]

We are getting to the point where Apple are the number one software publisher for OS X. I think that Apple uses the Altivec part of the G4 to good effect and the applications that are iLife have no equals in the other operating systems. Personally I am not affected by this perceived performance gap everyone keeps posting about.

I read this week that Motorola have the next generation G4 ready and that would be the safest bet to make for Apple hardware this year. I am not sure about this one but unlike the current G4s, this one can work in quad configurations. Just don't quote me on that.

It seems like the 970 has gotten a speed bump to 2.5 GHz without it being released yet! I have high hopes for this processor but it is next-gen hardware that will take time to come to market. 2004 is a good prediction but do not be surprised if it is the 2nd half of that year when Apple releases such computers.

[jethroted]

If this is true, then apple is in big trouble. By next year they should br considering weather or not to have their high end systems running the power 5 instead of the 970 running in their summer line up. At that point the 970 should be in the ibooks, and imacs. They should not be thinking of introducing the 970 at that point. Way too late. Of course that is mostly wishful thinking, as apples new way is "let's sell them old crap at a high price, and prtend speed doesn't matter." They have got to get ship shape real soon. We're getting tired of their games. The G4 was amazing when it came out, but that was so long ago. Moto has caused a lot of problems for apple, and those who defend/support them. Apple has to get with someone who can launch them far beyond the x86 landscape. Real soon.

[MacBandit]

Originally posted by Freg3000
I'd rather believe the Page 2 rumor that says it'll be available in July. :(

They're called Page2Rumors for a reason.

I am another one of those that all along have said we will not see completely new processor (i.e. PPC970) until MacWorld in SanFrancisco '04 at the earliest.

My reasoning is made up of a bunch of factors but the simplest is the fact that computer designs take any where from 8months to a year and a half for completion. Even if Apple had a PPC970 for test in December it would be October at the very earliest before they would have production of the finished computer.

[Dont Hurt Me]

Please dont get me wrong the g4 has been good ,but its increase in performance path has been horrible. The many people with recent powermacs say the past2-3 years have no real reason to upgrade since motorola has not gone anywhere. This is great for me and my machine but bad business for apple. Base powermac is a 1 giger and 133 bus. Well heck if i need the small bump again from motorola i can just choose a powerlogix,sonnet upgrade and again apple looses sales because they have not gone anywhere. even my 2 yr old 133 bus is current when you look at the base powermac though architecture has changed slightly. If Apple is to increase market share then they have to come out with a machine they says to me and everyone else that this is a must buy performance screamer. A 970 based Powermac could do this. Also the people that say all this engineering has to be done are wrong. They have been using the same case for years with minor revisions and considering that the current windtunnels are cooling 2 cpus it would be easy to stick in the 970 since it will be cooler and use less power then 2 G4s. Everything points to this chip now the question is when? If apple waits till next year then powermac sales will continue the decline. If apple breaks out the 970 this year then the sales are going to increase. Am I wrong here?

[MacBandit]

Originally posted by Dont Hurt Me
Please dont get me wrong the g4 has been good ,but its increase in performance path has been horrible. The many people with recent powermacs say the past2-3 years have no real reason to upgrade since motorola has not gone anywhere. This is great for me and my machine but bad business for apple. Base powermac is a 1 giger and 133 bus. Well heck if i need the small bump again from motorola i can just choose a powerlogix,sonnet upgrade and again apple looses sales because they have not gone anywhere. even my 2 yr old 133 bus is current when you look at the base powermac though architecture has changed slightly. If Apple is to increase market share then they have to come out with a machine they says to me and everyone else that this is a must buy performance screamer. A 970 based Powermac could do this. Also the people that say all this engineering has to be done are wrong. They have been using the same case for years with minor revisions and considering that the current windtunnels are cooling 2 cpus it would be easy to stick in the 970 since it will be cooler and use less power then 2 G4s. Everything points to this chip now the question is when? If apple waits till next year then powermac sales will continue the decline. If apple breaks out the 970 this year then the sales are going to increase. Am I wrong here?

Let's see new chip mean new Bus, new bus means new ram, new bus means new board, etc, etc, etc..

[Dont Hurt Me]

new mother board isnt going to take that long,new ram? ram is allready on the market working at those speeds on the pc side no issue. software has to be optimized,motherboard for chip is all thats needed maybe system controler? still what are you looking at? a new mother board and software thats it in a nutshell except for some more cosmetic changes on the case maybe a new color. And if this is to happen then they have probably been working on this since october or so.

[MacBandit]

Originally posted by Dont Hurt Me
new mother board isnt going to take that long,new ram? ram is allready on the market working at those speeds on the pc side no issue. software has to be optimized,motherboard for chip is all thats needed maybe system controler? still what are you looking at? a new mother board and software thats it in a nutshell except for some more cosmetic changes on the case maybe a new color. And if this is to happen then they have probably been working on this since october or so.

Software also has to be debugged and the whole design has to go through quality and user satisfaction testing.

[Brother Mugga]

Never posted here before; hi...

I've been watching the Mac site boards for about 9 months, waiting for the right moment to switch AND take about four or five of my students with me (in some kind of bizarre 'and then turned the floppy-drive on themselves' suicide pact).

[n.b. the only piece of Apple kit I have at the moment is this loverly 20" MY GOD IT'S SO BRIGHT I THINK MY RETINAE HAVE MELTED flatscreen that I'm looking at currently...through sunglasses, because I can't access the controls through XP].

During this time I've even managed to get a couple of friends to convert to iMacs because they weren't power-users and the new machines are actually good value. However...

I have about £3000 just sitting waiting for a decent machine from Apple so I can do graphics and animation. I don't care about games (the usual reason I'm unable to convert people), software compatability (I'll be building from ground up), or even the dearth of decent peripherals with OS X native drivers (tsch...). I DO care, however, about being lumbered with a machine that runs about half as fast on processor intensive tasks in comparison to similarly priced Wintels (and STILL doesn't have USB 2.0, further knackering its peripheral base - it's no good saying firewire is better; I agree with you. But that's of bugger-all relevance if no-one's making firewire devices).

If you check back on the boards, you'll see that the 'G5' has been anticipated for over a year. First it was going to be last summer, then end of last year, then middle of this year (with a switch from Motorolla to IBM). That's not good.

The problem of waiting for the new processor is further exacerbated by Apple's 6-month upgrade cycle. I know it might help them with product stability, technical support, and inventory management, it's REALLY knackering its switcher market. If a new Xeon comes out, Armari will wallop it into a (non-bottlenecked) workstation within about two weeks. And if they don't, I can always go somewhere else. Because there isn't a monopoly on Wintel harware.

I REALLY want to switch to Apple; I like the attention to detail in the hardware (FSB, recent noise problems, failure to site front panel Firewire/USB connections (etc.) notwithstanding). I love the OS. I like the idea of being part of a community, rather than just a target-market. I find you lot on the boards generally helpful, amusing, and surprisingly well mannered (although all that will change if and when I switch, of course). The trouble is, unless I see a 970 machine (or similar) arriving sometime around or before autumn, I will HAVE to stay Wintel; a wheezing, boiling 1.42+a-little-bit G4 just won't let me work efficiently. Which means that Apple will not only lose the money I and my student's would have spent on their machines, but also the concomitant benefits of an expanded user-base.

It therefore seems to me that Apple should do one of two things:

(1) Stick to a 6-monthly upgrade cycle but TELL US (roughly)what's coming so we can make an educated decision about when to spend (I understand the counter-arguments, but they only work if you don't trust the company not to stick you with soon-to-be-obselete gear or you think Steve Jobs is justified in lurching the inventory deployment cycle in order to 'wow' at MacWorlds)

OR

(2) By doing what the Wintel world does and upgrading regularly so you don't feel this dire need to time your purchase perfectly or get stuck with something that'll make you feel the company has given you a charming little marigold-moment (without lubrication).


Sorry for the long rant (although I clearly enjoyed it), it's just that I'm currently trying to get three or four students to hold off from investing in Wintel workstations on the premise that something competitive from Apple might be just about to arrive. And given the length of time this has been going on, I'm starting to wilt...

Brother Mugga

[Dont Hurt Me]

Macbandit you are right, but if Ibm is going to show prototypes to the public next month then i would bet apple is currently working on their stuff and putting the polish on it. If everything goes fine and from the sound of it (2.5ghz!) it is. Mass production this half and maybe on the shelves late summer?That still give those Superbrains in the software division time to get panther ready. Lets hope so.

[barkmonster]

wouldn't apple be shooting themselves in the foot by going straight for a 2.5Ghz chip in a model this year ?

If they went for a line up something along the lines of 1.4Ghz, dual 1.6Ghz and dual 2Ghz, they'd have more than enough room to bring out another range in january / february of 2004 at up to 2.5Ghz.

This would also give IBM around 18 months from now to come up with versions that are faster than 2.5Ghz.

I'm sure a boost of almost 600Mhz between now and the next range of macs AND a far more powerful cpu coupled with the kind of bandwidth that the G4 has been starved of would be more than enough for apple to wow plenty of people into buying a new powermac and you'd know there's a definate 25% speed boost coming 6 months later.

[technocoy]

I know that apple has signed on for the 970...

as well, even if it isn't until next year (january) how do you figure it should already be in iBooks and iMacs by then?? in case you are all forgetting, unless something changes in the x86 realm... apple is about to be the first to market with comsumer 64bit machines in mass production... they could release these at 1.2 ghz if they wanted... it's 64 bit, it's going to blow away the current proccessors on the market out of the water. And if it is the "year of the laptop", what's to stop them from letting the PM sales lag a little, and packing the 1.4 and such into the sweet new laptops... consumers are really wanting laptops right now. If you were apple, and you knew you were preparing to release the most powerful desktop on the market in 9mos, and that the only thing you could do about the desktops you have now is make incremental bumps with motorola processors, why not kick the **** out of the laptop market, and reclaim the throne as the fastest desktop maker next year? OS X has only to be recompiled for 64 bit, and when those desktops are released with 64 bit 970's all of you who run the mouth now will be flocking to buy one, whether it's 800 mhz 64 bit, or 2.5.... and last time i checked the whole Percieved speed gap between pc and mac is getting ready to crumble anyway when intel and such have to explain why Mhz isn't the most important thing as the tables are turned when they're releasing their 1 ghz 64 bit processor and apple is turning out their increasingly popular unix based ultra-stable non tight-assed operating system on a processor that's what? what was that? clocking higher than the new intel and AMD chips??!!! oh my GOD it going to be SWEET....

The world wants to move to mac... they just don't know it yet... but it's coming. it's coming!!! mwhaahaahaa!!!

technocoy:D

[nuckinfutz]

If they went for a line up something along the lines of 1.4Ghz, dual 1.6Ghz and dual 2Ghz, they'd have more than enough room to bring out another range in january / february of 2004 at up to 2.5Ghz.

I know there are battles going on within Apple right now on how to parlay these machines in such a way as to maximize profits. That's very risky however IMO.

I say it would have to be based on Yields. If IBM is getting sufficient yields of 2.5Ghz CPU's to satiate Apples demands then I say go for the gusto. After all this will be the only generation of PPC 970's on 130nanometer. They'll be ramping to 90nanometer quickly, perhaps allowing Apple to hit 3Ghz next year in 2H.

At any rate I think Apple still has much leeway if their able to get decent amounts of CPU's. I hope they do it right. It's time to start shipping some boxes!

[Dont Hurt Me]

I dont think we will see a 2.5ghz this year,nor do i expect a dual machine. 1.4, 1.6 1.8 sounds good to me AND a 1.2 or 1.4 imac 970. Then they can scale these up as they see fit since the cpu will allow a lot of wiggle room unlike the go nowhere motobutt g4. I still think all the duallies stuff was just trying to make up for where moto wasnt going. Hard to say what apple will do.

[cubist]

The Apple execs are worried about the 970 rumors hurting PowerMac sales. This puts them between a rock and a hard place: If they publicly say they are going to use the 970, that will definitely kill PM sales. If they say nothing, some people will assume nothing better is coming (especially after months of 1.0 then 1.2 then 1.4 dribbling updates) and buy PCs.

So I have a suggestion for Apple: Leak something. Not something absolutely conclusive, but something pretty persuasive. Give it a day or two, then send the usual recall missive. It won't hurt PM sales, because those who are hot on these rumors are waiting anyway. But it will give the rumor-hounds something to be excited about, courage to talk up the Mac at work and school, because the greatest Macs of all time are coming.

And as long as I have your attention, we need a better compiler. We hear that IBM has a highly optimizing compiler for the PPC. Please look into it. License it, and give it to everyone. Get Adobe and Microsoft to use it. It could make a huge difference.

[Dont Hurt Me]

i agree somewhat cubist, but i think what is really hurting powermac sales is motorola stagnation. If they want the sales then they will have to break from the g4 and do it as fast as possible and with a massive advertising campaign. After all they are wanting switchers and they are not going to get those in quantity as long as macs cost more and are still slower then the wintells.( the smarter ones know the OS is superior but the hardware is not)

[Brother Mugga]

r.e. Don't Hurt Me's post (the relevant 970 stuff is near the end, by the way. The rest is precursive).

Yes, I agree. The OS is superior (although XP isn't too bad). And yes, it is principally the cost:power ratio that puts many off. But it's really not the only thing that makes or breaks a switcher. This is why I have to chuckle when people say things like 'yeah, having a fast(er) chip is going to have people flooding over to Apple.' No it won't. It will help, but it's just part of a package of reasons. It's like they don't have any mates with Wintel machines or, if they do, they've never talked to them. The performance difference is important (see my previous post), but it's not the be all and end all that some Mac-a-holics think it is; they seem to be confusing their own frustrations with those of a switcher.

Most people I talk to ask the following questions:

1) Why does it cost so much?
2) Do I need new software?
3) Will my printer/scanner/etc. work with it?
4) Where do I get it fixed?
5) Can I play games with it?
6) WHO ELSE SELL'S 'EM?

I can kind of get round the first one (even though PowerMacs are still too expensive in my opinion). I can explain the benefits of OS X and hence get them over the hump of the second one. The third one is a git (see below), and it still amazes me at how few scanners and printers are OS X native when one of Apple's key markets is for graphics. Question 4 is trickier in Britain than the U.S. (new Applestores on the way though?). Question 5: Games? Schmames. Buy a console.
The real killer (and I don't think many Mac-users appreciate this) is the final question. Apple has an OS/hardware monopoly. Not even M$ has that (and imagine how indignant you'd all be if it did?). People don't like not having a choice of vendors; it makes us feel we're potentially getting ripped off (and, let's face it, if you look at some of Apple's business moves over the last 6 months, it does seem to be wringing every last drop of dosh out of its loyal customer base).

I don't see this talked about much on the boards, but it comes up a lot when I'm talking to fellow Wintels about Apple. They don't like the idea that EVERYTHING (pretty much) to do with their computer is going to be determined by the business plan and output of one company.

As a result, it's hard(er) to get people to switch, and so Apple's market share doesn't increase, and so there follows a paucity of OS X compatible peripherals and software (specifically games, although to be honest I don't give monkies about that) and hence the problem trying to get people to switch continues...and so the cycle persists.

So, am I hoping for the 970 soon? Yes, or I won't switch. But one of the main reasons for this will be that if Apple doesn't shift things along, no-one else will, because it's a monopoly. And I don't wan't to invest in a (potentially) sinking ship. This is why the idea that 970s might not arrive till next year is pushing me way over in the direction of Dual Xeon Armari. In fact, the website is loaded as we speak. Of course, if Apple could confirm that the 970s will be out in a few months...

I understand that the 'mac clone' era almost killed Apple, but this just seems to indicate that Apple was bloated and unable to compete effectively...as it often appears now. I think they might need some competition on the hardware side; after all, they TELL us theirs is the best, but won't let anyone else put that to the test. Without a hardware competitor, I think they may just keep milking their customer base with a grind up through even more tiny G4 'boosts' and then a slow roll-out of 970 (I guess we'll know one way or another by the end of the year). BUT, oh, hang on, if a PC company tried to do that, it would die like a dog, because all the competition would pounce on the latest and/or best value chips and wander off with their customers.

Let's face it, we all have a chuckle at Gateway...but if Apple customers (with a lot of investment in the OS) had been free to go to an alternative vendor over the last few years then Apple could well have gone the same way...

Brother Mugga

PS: As you can see from the length of these posts, time is hanging heavy until The Premiership at 10.30...

[macrumors12345]

Originally posted by nuckinfutz
I say it would have to be based on Yields. If IBM is getting sufficient yields of 2.5Ghz CPU's to satiate Apples demands then I say go for the gusto. After all this will be the only generation of PPC 970's on 130nanometer. They'll be ramping to 90nanometer quickly, perhaps allowing Apple to hit 3Ghz next year in 2H.


Unfortunately, if you look at the ZDNet article that started this thread, you will see that the implication is that it is the .09 micron version that will run at 2.5 Ghz, not the .13 micron version.

I too was extremely impressed when the IBM Press Release said that the .13 micron 970 would scale up to 2.5 Ghz. It would mean that the 970 would significantly outperform the Pentium 4 on an equivalent process, since the P4 will only reach about 3.2 Ghz or so on the .13 micron process (and a 3.2 Ghz P4 would be no match for a 2.5 Ghz 970 in integer performance, let alone fp or vector performance - and all at lower power consumption no less).

Of course, maybe the ZDNet article is wrong (given that it is generally a cut-and-paste job) and the IBM Press Release was correct. I really hope that is true. But the fact that the IBM Press Release has now been pulled from their website certainly doesn't look promising - it sounds like somebody in the PR department might have made a serious mistake in implying that the .13 micron 970 will scale to 2.5 Ghz. Sigh. Of course, even a 1.8 Ghz 970 will be much better than what we have now, and for the first time in several years we will actually be competitive with Intel/AMD offerings (on things besides vector operations). But a 2.5 Ghz 970 really would have leap-frogged the Pentium 4 and Athlon-64 by a good margin...well, it's nice to dream!

[Brother Mugga]

Yeah, I just presumed that the ZNet was a cut'n paste cock-up because otherwise...well, what are we (and specifically Macrumors) getting excited about?

I didn't realise it'd now been pulled by IBM. That doesn't look good. Any chance of arn confirming whether it's meant to be 1,8-2.5 at o.13 or whether the 1.8+ relates just to the lower process?

Brother Mugga

[zac4mac]

I find that idea a lot less likely than Steve was on the phone screaming that he has enough trouble with these 970 rumors killing PM sales and IBM pulled it as a nicety. See if they still show the protos at CEBIT or Apple makes an announcement before that. Hoping for a summer release - I want a rev 2 or 3 next year.

Z

[Sol]

To the teacher telling his students to wait for Apple: do not hold your breaths. Apple takes its sweet time with upgrades and if their current line-up is not good enough for you now then it will never be good enough.

Those who say things like "my $5 Dell blows away the top of the line PowerMac AND it has USB 2.0" are probably comparing how many frames per second Quake 3 can render on the two platforms. The G4 processor may be underutilized by most programmers but when Altivec and dual processors are taken into consideration then it is the Pentium chips that are left sucking eggs.

As for the architecture differences, you are comparing apples to oranges. Apple has never taken the bigger is better approach to hardware design and if you cannot live with that then too bad. Personally I find value in a computer that is optimized to be powerful and stable.

Several months ago Apple brought out the first consumer machines with Silicon On Insulator CPUs. I do not remember which version of the G4 that was but I do remember that Apple was the first. Funny how no-one mentions innovations like this when the focus is on increasing GHz. Anyway, say what you will about the G4 but the fact is that without it Apple would not be an industry leader when it comes to fan-less and portable designs.

The theory that rumor sites are affecting the sales of PowerMacs may be partially true but I think that misguided Apple users are also to blame. They are the first to suggest that Intel hardware is better while easily editing movies, burning DVDs, etc that their "slower" Apple hardware and software lets them do. Any potential switchers reading comments from these misguided Apple users probably nod their heads in agreement and return to Wintel in the hope that GHz will somehow empower them. Too bad the software on Windows is so confusing.

When Quark XPress finally gets released for OS X one of the last hurdles will be overcome. Unfortunately a lot of you will jump on the "Where are the G5s" bandwagon and continue spreading this negativity about Apple products. The rest of us will surely feel a degree of pride in being part of the Mac club.

[Dont Hurt Me]

Nice post sol but apple hardware is not up to speed with the wintel stuff and this is fact. Look at buses,ddr ram that isnt fully used and speed, look at benchmarks on cpu tests,look at any benchmarks for photo apps or games or you can look at some flash benchmarks people were doing today on this site. Bottom line is the hardware is behind or slower if you will. Software the mac is superior, iapps,The OS,ease of use stability etc. They have been holding on to the G4 for to long and its showing. I love my Mac but if they are going to charge top dollar and are supposed to be the cream of the crop then the hardware needs to catch up to the software. Lil Motobumps just aint getting the job done hence 3% marketshare. Rumor sites arent affecting sales, Slow minor increases with motorolas stagnating g4s are!

[Sol]

Whatever benchmarks you are referring to I can tell you that there would be other benchmarks showing the Apple hardware as the superior machines. I think someone had links for such benchmarks earlier in this thread. I also refer you to the speed comparisons performed by Steve Jobs with Photoshop. As for the games, there are many reasons why frames are lower on OS X than on Windows and most of those are software based. Remember that most games for OS X are ports from Windows so there is bound to be a generation loss when it comes to performance.

You say that Apple has held on to the G4 for too long but it can also be argued that Windows has held on to the Pentium for too long. It would be no exaggeration to say that the x86 architecture is hopelessly outdated and the only reason why it is still developed is that the domination of Microsoft's operating systems has created such a huge demand for it.

There are benefits to PPC RISC chips and chief amongst those is the heat efficiency. Consumer RISC processors are also set to go 64 bit this year while Intel has dismissed the need for such progress. Perhaps it would clash with their dependance on Windows.

In any case, where you see bottlenecks in Apple's hardware I see efficiency and superior design. If bigger is better for you then there is no point in spending your money with Apple. If you can appreciate things that are well-built then by all means, pay the premium price and get a premium product.

[Brother Mugga]

Sol,

I wouldn't have thought many would dispute that PPC architecture is superior to x86; it's one of the reasons I want to switch (although I feel it IS something of an exaggeration to suggest that the latter is 'hopelessly outdated'...and on the RISC/CISC front most chips are effectively lopsided hybrids now anyway).

But, hang on, wasn't Betamax was superior to VHS? (etc.)

Similarly, I agree that if apps. were cocoa and coded efficiently for altivec then the G4 could compete...but they aren't, so it doesn't. Your line of argument exhibits the problem that I've been alluding to in all my posts; that Mac-users tend to talk about potential rather than living in the here and now. For example, note all the posts over the last year talking about the 'G5' (or whatever) in which posters say 'how cool would that be! We'll really hammer (no pun intended...well, okay, maybe a little) AMD and Intel and people will switch in droves' (etc.). Meanwhile, the Athlon and P4 have steadily scaled to speeds that - sorry, but as Dont Hurt Me points out, in any 'real world' test - leave the G4 puffing along like an asthmatic lard-arse.

I fear the same might happen with 970s 64-bit possibilites; don't forget that re-coding will be needed to take advantage of this and how many major software houses are even bothering to write in cocoa, rather than carbon? Be prepared for more 'potential over real world' board arguments.

It's no good moaning that it's only the installed user base of Windows that is driving this process; I agree. But the fact remains that it's happening (and I wouldn't underate the effects of competition either (see previous post) - see what happened to Intel's R&D when AMD started making inroads. Expect similar between Nvidia and ATi now the latter is back in the saddle).

Finally (in response to your final paragraph), it's NOT that bigger is better per se, but rather that there comes a point when chip speed really does matter for some applications. I agree with you; I really wouldn't mind paying a premium for a premium product. That's why I will switch if it arrives soon enough. Because these current PowerMacs AREN'T premium. XP is (generally) stable enough (not that I wouldn't prefer OS X) and I can get workstations (that do not have bottlenecks...I'm not quite sure how you percieve a FSB that chokes the processer as evidence of the Mac's 'efficiency and superior design'?) that are pretty well designed and really fly (although I'd prefer a Mac that COMPETED, and wouldn't mind paying a premium).

People often come out with the 'comparing Wintel with Apple is like comparing Ford with BMW' analogy. Yeah, to an extent. But how long do you think BMW would last if they only managed to put 1.1 engines (chip) in them and you couldn't go on half the roads in the country because of the suspension or whatever (OS), whereas Fords were knocking out 2 litres (for half the cost) that could go pretty much anywhere?

Now do you see why I fear for Apple unless they start whacking out competitive processors (and soon)?

Brother Mugga

PS: Regarding heat, Google up 'Centrino' and also 'Prescott' for Intel. Not 970s, but a lot better than the P4.

PPS: Notice also that, in talking up the Centrino Intel are having to shaft the 'megahertz myth' themselves (just as with the Itanium, but now for a CONSUMER chip). Now, come on Sol, that's GOT to have you laughing almost as much as me...?

[Sol]

Now that you mentioned it, yes it does seem ironic that Intel will try to convince its customers that lower MHz does not mean less performance. A few weeks ago Intel announced that its next-gen mobile processors will come out at something like 1.6 -1.8 GHz speeds in the middle of the year. It occurred to me then that (in the mobile space) Apple will suddenly close the MHz gap all thanks to Intel.

Anyway, the applications that matter on OS X are those made by Apple, Adobe, Macromedia and Quark. Apple's iLife suite has no equals in other OSs and I guess it is due to the optimizations for Altivec and dual CPUs. Adobe has a reputation for taking advantage of new hardware features and Photoshop is a prime example of 3rd party software done right. Macromedia continues to provide the goods and have supported OS X from the beginning. Quark are... Well, the less said about them the better. Major let-down for what is supposed to be an industry leader in layout software. My point is that while the average programmer does not optimize for Apple hardware, the ones that matter do. Except for Quack, I mean Quark.

A point about Carbon: in retrospect I think Apple made a mistake by encouraging developers to Carbonize their applications. The Classic layer all-ready gave us compatibility with older applications. If Carbon was not part of the equation then chances are that developers would have been forced to re-write their work in Cocoa. Surely it would have taken longer to see an OS X version of Photoshop or even Final Cut Pro but in the meantime OS 9 would have been good enough for most people.

[barkmonster]

In any case, where you see bottlenecks in Apple's hardware I see efficiency and superior design. If bigger is better for you then there is no point in spending your money with Apple. If you can appreciate things that are well-built then by all means, pay the premium price and get a premium product.

Do you think Apple might release a 'premium product', that can match a home built PC from about 18 months to 2 years ago sometime soon ?

Superior design that includes motherboards that render a whole range of powermacs next to useless for audio and a cooling system that was obviously inspired by a hairdryer isn't exactly something to brag about. The TiBook, iBook, iPod etc... are all examples of superior design but there's really not much that would make me do anything other than buy a cpu upgrade for my current mac right now.

For the software I use mainly (Protools LE) there's a few things that kind of piss me off right now. The positive things are that Protools 6.0 is out and it takes full advantage of both cpus on a dual system and runs under OS X. The negative thing is the performance compared with typical PCs and even older macs.

The current dual 1.25GHz model is under 60% faster than a properly configured dual 500Mhz sawtooth model. Where's the progress ?

The same dual 1.25Ghz model, even though it's running under OS X and the software is fully multithreaded so both cpus are being used to their full potential, is STILL around 20% slower than a custom configured Athlon PC that can be built for under $1000!

That's not even the fastest PC system, some people with very recent "thoroughbred" Athlons and 2.4Ghz+ Pentium 4s are running well over 60% more plug-ins that the mac can manage.

I wouldn't pay full price for a brand new mac (if I had the cash) until there's some genuine power, I want a mac that's going to last me a few years, if it's already dated THIS year, why would I bother buying one ?

I could get a used digital audio G4, pop in one of those powerlogix dual upgrades and max out the RAM for about the same price as the entry level 1Ghz G4. I could also actually "use" the thing too, It's a long up hill struggle of waiting for software synths and plug-ins to be released for OS X. I don't have boatloads of outboard effects, samplers, synths etc... I have a mac, protools LE, a few plug-ins and a few software synths. It's running on an old mac, I've been waiting years for at least an entry level 1Ghz G4 before I give up on my beige G3, now it's here, it's OS X only and I can't use it!

Seeing as whatever the CPU situation on the mac, I'm waiting for software to be ported to OS X, by the time there's loads of it available, the PPC970 based macs could be out. That's worth waiting for.

It's got all the bandwidth the G4 lacks, that's half the reason the AMD and Intel based PCs are slaughtering the G4 in certain benchmarks, 2.1 - 4.2Gb/s is a lot better than the 1.3Gb/s the G4s are limited to.

[Brother Mugga]

I'm inclined to agree with you, Sol. I think encouraging carbonisation just tied up coders who could have been working on cocoa. Mind you, I'm not exactly au fait with how software houses go about deciding these things, so that's just an uneducated opinion on my part.

r.e. the Centrino/Powerbook stuff - absolutely. Suddenly that G4 1 Gig in the Powerbook is going to start looking much more attractive to the big-number=big-value monkies around the world (indeed, even considered PC opinion (i.e. PC Advisor magazine) is hailing them (rightly) as class acts). This is probably why Mr. Jobs has labelled this year the 'year of the notebook/laptop' - because the power consumption to performance ratio of the G4 allows it to compete more effectively in this market.

My worry about this is that Jobs' statement perhaps implies that the 970 won't make it into Powermacs this year...which means Apple is probably going to lose me and my mates/students as we need to get workstations by late summer/early autumn. Pah.

Brother Mugga

PS: Must go now - England are just giving us all hope in the cricket (before no doubt rudely dashing it as Australia power home with 5 overs to spare).

PPS: Update...

Sorry to doubt you, England. What I meant was keep us all on tenderhooks and then screw it up in the PENULTIMATE over.

I despair...

(Sorry if this makes no sense to any Americans out there. It's a cricket world cup thing).

[Dont Hurt Me]

Nice post barkmonster, the hardware is behind this is no secret. this why they are using 2 cpu's -1.25 &1.25 does not equal 2.50 no matter how hard you try. The whole computing experience is better on the mac period. I own 3 macs right now but the bottom line is that the hardware is on the back row on the race track and if they want to be the pole sitter then they are going to need the 970 as soon as they can because soon these g4's powermacs wont be on the race track. Iam not trying to bust on the mac at all, iam just stating the facts that the hardware division needs to let go of the g4. The g4 has held back the Mac for to long. I think most of us know this and is why there is so much interest in where cpu's and the mac are headed.

[Coca-Cola]

Follow the money. I think that would be wise to do in this case. Intel has the fastest chips because they have all the money. Money for development and research. Motorola has money but must play it safe for now. The safe play is cell phones. Who has even more money then either one of them? IBM. IBM is the mother load of research money. They will have the fastest chip.

[macrumors12345]

Originally posted by zac4mac
I find that idea a lot less likely than Steve was on the phone screaming that he has enough trouble with these 970 rumors killing PM sales and IBM pulled it as a nicety.Z

Conceivably, yes. But it does not help your case (which, trust me, I want to believe, but I am being realistic here) that the ZDNet story linked at the beginning of this thread specifically says that the 2.5 Ghz version is the *next* release of the 970, not the first release of the 970.

[whooleytoo]

I'm surprised some are saying the reasons the G4 is so slow is simply poor Altivec support in most apps, the simple truth is not all applications are at all suited to vector processing. For data intensive tasks it is suited to (eg. encryption/decryption) it offers huge performance boosts, but for many (most?) apps it isn't that useful.

By the way, how many people remember a computer company called the Osbourne? They had a hugely successful computer called the Osbourne I many years ago, were getting rave reviews in the media, selling like hotcakes, and had warehouses full with people screaming to get their hands on them. Then, someone thought it would be a great idea to preannounce the Osbourne II - a much improved model. Sales of the Osbourne I stopped dead. They couldn't clear out the existing inventory, and without a revenue stream, couldn't complete the Osbourne II. Bye bye Osbourne.

Apple had better *not* preannounce the 970 machines (well, apart from the announce now/shipping in 2 months kinda thing..:-)

One other thing, I hope Apple doesn't ignore the gaming market, it's one of the few things that entices people to upgrade their machines more often - a crucial area since there aren't many first time buyers in the 'developed' markets any more!

Mike.

[Brother Mugga]

I can see the argument (r.e. ye olde Osbourne effect) and it perhaps has some validity, if only because of the stupid way Apple manage their upgrades.

When you're thinking of buying a PC you know there'll be something better along in a month, but because the improvements are (generally) frequent and incrementally small you still go for it. As I noted previously, this process is driven principally by the way in which competition affects the marketplace (i.e. a choice of chips, OEMs, and vendors leads to rapid development and deployment of hardware, lest people go elsewhere).

Contrast this with Apple's 6-monthly upgrade cycle, with might have been incrementally small of late, but is clearly not frequent. As a consequence, the only sensible time to buy is near the beginning of the cycle (unless you're desperate for a machine at work - see below). The relevance of the Osbourne effect has also been increased by the slow scaling of the G4, which has resulted in Mac users looking for a huge leap in performance to catch up with the Wintel world.

However, despite this, there seem to be other, perhaps more persuasive reasons for Apple actually coming clean about the 970.

Powermac sales have already stagnated for want of juice (it's one of the reasons Apple is looking to wring every last drop out of other revenue streams). Furthermore, I know companies that are currently using Macs that are thinking of switching to Wintel because (Quark notwithstanding), with the switch to OS X, they're going to have to upgrade their software anyway, so why stick with a platform that seems to be spiralling away from the power-user? This is a great shame as, unlike x86/XP, PowerPC/OS X really can claim to be 'next gen'. If Apple just came out and said, 'Yes, we're going 970 before the end of the year' I really think it would help, rather than hinder sales (certainly in the long run). The majority of those purchasing Powermacs currently are people/companies that really need them (and now) for their work, so trailing the 970 wouldn't affect these sales: indeed, some of those thinking of switching to Wintel might actually hold off because they would have been reassured that they could still have a bright future with Apple.

It is tricky, but, with this in mind, I'm not convinced the Osbourne effect applies. Certainly, as a potential switcher, if I don't at least hear something uplifting from Apple fairly soon then I will HAVE to stick with Wintel as I'll need a new workstation before long. Knowing that the 970 is coming out by August would encourage me to hang on for an extra couple of months (even if it hurt a bit) and gain Apple some new converts. In fact, it's only by talking up the 970 that I've managed to keep some of my friends interested in the possiblity of getting Macs.

Hang on, why aren't I on commission? I don't even own a mac yet, for god's sake...

Good points about Altivec and gaming, by the way.

Brother Mugga

PS: Hopefully with IBM pushing things along Apple could get into a bi- or tri-monthly upgrade cycle (rather than the 6-7 currently) as well. Here's hoping...