PDA

View Full Version : As if the Pentium M weren't already way too fast for the G4...


spaceballl
Sep 10, 2004, 04:38 PM
http://www.theinquirer.net/?article=18329
Dual Core Pentium Ms... to save power, it can turn a whole core off. Let's hope that Motorola can compete with this with their dual core CPUs... or hopefully the G5 will have flourished by then and be ready for notebook time.
A Dual Core Pentium M is... REALLY fast.
-Kevin

mikeyredk
Sep 10, 2004, 04:50 PM
now its switched the desktop processors rock while the notebook processors stink

olden times it was the other way around. lets hope freesemi doesn't choke like its big brother.

SiliconAddict
Sep 11, 2004, 01:07 AM
Forget about Moto competing. I'm more interested in seeing if they can even release a product in a timely manner. Everyone is drooling at Moto's dual core dreams and seems to forget how badly they have hurt Apple over the years with delays and subpar hardware.

I won't believe a word out of that company until Jobs himself announces it on stage.

QCassidy352
Sep 11, 2004, 01:44 AM
um... wouldn't a dual core pentium M be faster than desktop PCs? The top end (1.7 Ghz) pentium Ms already outperform the notebooks with 3 Ghz P4s. Desktop PCs are only up to what, 3.4, 3.6 Ghz or something?

btw, anyone see that these will use a 65 nm process? So much for IBM's theory that the entire industry was having trouble at 90 nm... :(

EminenceGrise
Sep 11, 2004, 03:09 AM
I won't believe a word out of that company until Jobs himself announces it on stage.

What, you mean like "3.0 GHz [G5's] in a year"? Please. I'm not going to argue that Moto hasn't hosed things in recent history, but they do appear to be on the right track by spinning off the chip division as Freescale, etc. Don't think that just because G4 development languished under Moto's more recent abysmal management (which is now no longer directly responsible for Freescale, BTW), that means they can't still hit one out of the park. Particularly now that they face losing a shedload of business to IBM's G5. My guess is that there is a good chance we'll see Frescales new 'G6' chip in what everyone for some inexplicable reason wants to be the mythical 'Powerbook G5'. Particularly if they have perfected their GaAs process that they reportedly had figured out how to do cheaply a couple years back, though they don't even need that. Clock for clock the G4 is faster than the G5 (esp. for Altivec instructions), so they need only bump the clock speed and bus speed. Even Moto, for all their problems, will eventually get this figured out.

As for Moto's "subpar" hardware and how they have "hurt Apple over the years", we seem to have a short memory. The G3, when released as "up to twice as fast" as the PII, yep big disaster for Apple there. The G4, "guarded by tanks" personal supercomputer vs. the "harmless" Pentium - yes, complete chaos for Apple. Moto has almost certainly helped Apple more than hurt them over the years, and these past 3-4 aren't much more than a blip in Apple's 1/4 century history. I doubt that Moto's company motto is "Let's hose Apple", so can we stop with the Moto bashing please. It's tiresome and accomplishes nothing, especially at this point.

spaceballl
Sep 11, 2004, 03:39 AM
btw, anyone see that these will use a 65 nm process? So much for IBM's theory that the entire industry was having trouble at 90 nm... :(
This isn't just IBM's theory. This is fact. But by trouble, they don't mean inability to produce. They just mean that when they hit 90nm, the historic rules of power reduction, less heat, etc, didn't really follow suit. That's why Intel isn't ramping their P4s up half as quickly as they were in the sub 2.5 ghz era of P4s. That's why we don't have 3 ghz. So this doesn't necessarily effect the transition to 65nm. The fact is though that Intel is nearing that stage. Let's hope IBM is too.

aptmunich
Sep 11, 2004, 06:40 AM
I think the problem is not making chips that small, the problem is making a million of them.

If every 4th chip you make breaks then the price per chip becomes unrealistic.

I guess that is the problem: consistently producing these hard to make chips at 90nm

Logically though, once they have all the wrinkles ironed out, they should be able to start tackling the 65nm hurdle at ibm as well (they probably already have...)

Abstract
Sep 11, 2004, 07:23 AM
Clock for clock the G4 is faster than the G5 (esp. for Altivec instructions), so they need only bump the clock speed and bus speed. Even Moto, for all their problems, will eventually get this figured out.

As for Moto's "subpar" hardware and how they have "hurt Apple over the years", we seem to have a short memory. The G3, when released as "up to twice as fast" as the PII, yep big disaster for Apple there. The G4, "guarded by tanks" personal supercomputer vs. the "harmless" Pentium - yes, complete chaos for Apple.

Firstly, not sure if that first bit about G5s vs G4's is true.

Secondly, IBM made the G3, not Moto. Motorola made the G4 used in Apple computers. The G3 was by IBM, and was never the problem.

On that note, I want to say that i generally agree with you, and would rather see a dual-core Freescale G4 than a G5 in a Powerbook. If they prove to be as fast as a G5 at the same clockspeed, then who cares if it's the G5 or not? Its marketing. I want a fast chip in the Powerbook, and if it's not IBM's G5, and is a Freescale chip instead, then fine. People just buy way too much into G5 marketing. If Apple offered something that's almost as fast, I'm sure there would be a complete Doughnut out there who'd hold out for a G5, "just 'cuz the Freescale chip is not a G5!" :rolleyes:

Mord
Sep 11, 2004, 01:14 PM
both ibm and motorola made g3's the ibm ones are way better though, there copper based cpu's compared to the aluminum based motorola ones

jhu
Sep 12, 2004, 06:20 PM
i'm pretty sure the early ppc750 computers, such as beige g3 or b&w g3, were motorola whereas the later ones, such as ibook, were made by ibm.

Celeron
Sep 12, 2004, 07:06 PM
Clearly most people agree that the Pentium M is leaps and bounds faster than the G4 processor. The question one has to ask is why Apple is still holding onto the G4 technology? Why not drop a PentiumM in the Powerbooks? They could still make them sleek and Apple looking in outward apperance and give the entire line quite a performance boost.

Given that all the hardware within the Powerbook is the same as a regular Intel-based laptop I don't see Apple's desire to remain with the G4. Just cut a version of OSX that only runs on the Powerbook's Intel hardware and be done with it.

That right there is the key to Apple's continued success. How much longer till we see a Powerbook G5? Who knows...

jhu
Sep 12, 2004, 08:59 PM
Clearly most people agree that the Pentium M is leaps and bounds faster than the G4 processor. The question one has to ask is why Apple is still holding onto the G4 technology? Why not drop a PentiumM in the Powerbooks? They could still make them sleek and Apple looking in outward apperance and give the entire line quite a performance boost.

Given that all the hardware within the Powerbook is the same as a regular Intel-based laptop I don't see Apple's desire to remain with the G4. Just cut a version of OSX that only runs on the Powerbook's Intel hardware and be done with it.

That right there is the key to Apple's continued success. How much longer till we see a Powerbook G5? Who knows...

the problem is: what software is going to be running on such a machine? emulation is out of the question since that would be slower than actually having a g4 in there. and no one's going to be recompiling their software just to run on a pentium m os x laptop.

MorganX
Sep 12, 2004, 09:06 PM
btw, anyone see that these will use a 65 nm process? So much for IBM's theory that the entire industry was having trouble at 90 nm... :(

Actually Intel says it can build cpus cooler at 65nm than 90nm, so IBM was correct. At least with regards to Pentiums. You don't see Prescott steppings getting much cooler. Centrino is a different story, but you don't see Intel pushing the 2Ghz version very hard either. It smokes all desktop Pentiums 4s and AMD FXs currently available. Well, maybe not smoke but beats. You can get one in a Sony laptop by adding $1000 to the price tag.

Intel is keeping the price high and availability low to push the Prescott it has invested so heavily in. Stinks.

wide
Sep 12, 2004, 09:10 PM
um... wouldn't a dual core pentium M be faster than desktop PCs? The top end (1.7 Ghz) pentium Ms already outperform the notebooks with 3 Ghz P4s. Desktop PCs are only up to what, 3.4, 3.6 Ghz or something?

Actually, the top end Pentium-Ms are at 2.0 GHz now with up to 2 MB cache (I think). You are correct about these being more powerful than desktop processors, which is why Intel is beginning a transition into the Pentium-M chipset from the Pentium 4.

jhu
Sep 12, 2004, 09:12 PM
Actually Intel says it can build cpus cooler at 65nm than 90nm, so IBM was correct. At least with regards to Pentiums. You don't see Prescott steppings getting much cooler. Centrino is a different story, but you don't see Intel pushing the 2Ghz version very hard either. It smokes all desktop Pentiums 4s and AMD FXs currently available. Well, maybe not smoke but beats. You can get one in a Sony laptop by adding $1000 to the price tag.

Intel is keeping the price high and availability low to push the Prescott it has invested so heavily in. Stinks.

if by "smokes" you mean the power requirements are lower or performance/power ratio, then yes. if instead you mean it has better raw integer/fpu performance, then no.

SiliconAddict
Sep 15, 2004, 11:02 PM
What, you mean like "3.0 GHz [G5's] in a year"? Please.

Yes. Please. Jobs made that claim. IBM didn't. And the fact of the matter is the reason they didn't hit 3Ghz is because of the trans to the 90nm process. You can bet your CPU that if they had the process down pat in spring we would be on a 3Ghz G5 right now. I'm willing to bet hard cash that IBM was cranking out 3Ghz G5's but the yeilds of stable 3Ghz G5's were so low that they scaled them back to 2.5Ghz.

Even Moto, for all their problems, will eventually get this figured out.

Yes. The key word of the day? EVENTUALLY. What does E=? 6 months? 1 year? 2 years?

As for Moto's "subpar" hardware and how they have "hurt Apple over the years", we seem to have a short memory. The G3, when released as "up to twice as fast" as the PII, yep big disaster for Apple there. The G4, "guarded by tanks" personal supercomputer vs. the "harmless" Pentium - yes, complete chaos for Apple. Moto has almost certainly helped Apple more than hurt them over the years, and these past 3-4 aren't much more than a blip in Apple's 1/4 century history. I doubt that Moto's company motto is "Let's hose Apple", so can we stop with the Moto bashing please. It's tiresome and accomplishes nothing, especially at this point.

Nope. Iím sorry but Appleís the one who has been hosed. The G3 was a strong CPU. So was the G4 in the beginning. But at the end moto unzipped their pants and hosed on them alright. With craptastic micro speedbumps that lagged farther and farther behind the competition. The G5 was and is a breath of fresh air. And as for a short memory. While its never been proven you seem to forget that last falls release of a speedbumped POwerbook was held up for unspecified reasons until after the start of the school season. It was reported on multiple sites that Jobs was foaming at the mouth because of it. Can anyone verify this? Nope because this is Apple after all. But yet again Apple gets bent over for the moto ride. Has Moto had some strong releases? Sure. But lets be real. Moto canít be trusted with their CPU roadmaps.

DrBoar
Sep 16, 2004, 08:53 AM
The B&W in the spring of 1999 had not only a top end 450 MHz G3, but also a very fast 100 MHz bus and the killer graphical card ATI 128. All 3 things well on par if not better than Wintel boxes, and then USB and FW as icing on the cake. :) Those G3s were made by IBM as was the previous top of the line 350 MHz 604E.

Five years ago Motorolas miscarriage named G4 slid into Apple computers. The first year was horrible with the 500 MHz wall and what is even more depressing is that during the following 4 years the G4 did nothing to close the MHz and performance gap to the Intel and AMDs. :mad:

If Freescale is going to make good CPUs they have a long track record to break free from. The last great CpU they made was the 40 MHz 68030 used in the IIfx. :rolleyes:

dieselg4
Sep 16, 2004, 09:17 AM
the problem is: what software is going to be running on such a machine? emulation is out of the question since that would be slower than actually having a g4 in there. and no one's going to be recompiling their software just to run on a pentium m os x laptop.

This sounds like a job for Transmeta, except their chips are really slow . . .

slughead
Sep 16, 2004, 02:19 PM
Yes. Please. Jobs made that claim. IBM didn't. And the fact of the matter is the reason they didn't hit 3Ghz is because of the trans to the 90nm process. You can bet your CPU that if they had the process down pat in spring we would be on a 3Ghz G5 right now. I'm willing to bet hard cash that IBM was cranking out 3Ghz G5's but the yeilds of stable 3Ghz G5's were so low that they scaled them back to 2.5Ghz.

I'd imagine this is true, along with the other problems we saw with the 2.5's (radiator production problems, etc).

Yes. The key word of the day? EVENTUALLY. What does E=? 6 months? 1 year? 2 years?

Yes, this is what I see with Intel's 65nm promise: Thank you for telling us your road map, now go back to sucking.

Everyone is headed to dual core and 65nm, including IBM and AMD. Intel's just talking loudly about it because AMD's been outselling them in some markets recently. One of the problems I see is that the market looks at Intel to be the market leader, when we know the Itanic was worthless, IBM got the first 64bit PC CPU, and AMD is still beating the pants off Intel for the race to the 64bit transition.

Intel made a big mistake IMO in overlooking x86-64, and they're paying for it now.

Nope. Iím sorry but Appleís the one who has been hosed. The G3 was a strong CPU. So was the G4 in the beginning. But at the end moto unzipped their pants and hosed on them alright. With craptastic micro speedbumps that lagged farther and farther behind the competition. The G5 was and is a breath of fresh air. And as for a short memory. While its never been proven you seem to forget that last falls release of a speedbumped POwerbook was held up for unspecified reasons until after the start of the school season. It was reported on multiple sites that Jobs was foaming at the mouth because of it. Can anyone verify this? Nope because this is Apple after all. But yet again Apple gets bent over for the moto ride. Has Moto had some strong releases? Sure. But lets be real. Moto canít be trusted with their CPU roadmaps.

I agree, don't let those "eats pentiums for lunch" ads (I still have a powerbook poster with that :D) fool you, the G3 may have been good in mid-98 but they were basically frozen in time until the G4s, which were great.. until they stopped advancing.

Now we have IBM's G5s and they're advancing at least comparable to the rest of the market (even with their 3ghz shortfall), as well as been a cut above immediately after coming out.

jdechko
Sep 16, 2004, 06:17 PM
I wonder if the naming convention of Intel is going to change... First we had the pentium, then PII, P3, P4 and now Pentium-M. Its interesting that we'd have thought that the next (logical) step would be a P5 followed by a P6. But currently, the Intel chip Processor architecture is the P6 - and it used to be P5. Im just thinking that Intel has decided to abandon sequential part numbering to avoid saying which one is "better" Windows has moved away from their year-number based system (yeah, i know, there's 2003, but thats server). Adobe has moved from numbered naming systems and has now adopted Adobe XXX CS to keep their stuff in line. I just see that as what may be happening in this case

jhu
Sep 18, 2004, 11:18 AM
they should call their next processor "sexium" because it's sexy and sex sells. oh yeah, sex...

JFreak
Sep 18, 2004, 11:52 AM
Clearly most people agree that the Pentium M is leaps and bounds faster than the G4 processor. The question one has to ask is why Apple is still holding onto the G4 technology? Why not drop a PentiumM in the Powerbooks?

you clearly know nothing or very little about technology; there's a reason why apple uses PowerPC processors and the consumer PC manufacturers use x86 processors - and that is called "binary compatibility".

you just cannot run PPC binaries on a x86 platform without any emulation, and i can assure you any kind of (hardware or software) emulation will give such a drastic performance hit that you just cannot even think about it as an option.

if apple would switch to x86 hardware, or even began to support it, that would mean that all macintosh developers would have to re-code their software; which would effectively mean either abandoning macintosh or not supporting x86 hardware - so, you see, there's no point for apple in doing something like that.

titaniumducky
Sep 18, 2004, 01:14 PM
Forget about Moto competing. I'm more interested in seeing if they can even release a product in a timely manner. Everyone is drooling at Moto's dual core dreams and seems to forget how badly they have hurt Apple over the years with delays and subpar hardware.

I won't believe a word out of that company until Jobs himself announces it on stage.

Have you not heard about IBM at all? It's becoming another Motorola.

Celeron
Sep 18, 2004, 01:37 PM
you clearly know nothing or very little about technology; there's a reason why apple uses PowerPC processors and the consumer PC manufacturers use x86 processors - and that is called "binary compatibility".

you just cannot run PPC binaries on a x86 platform without any emulation, and i can assure you any kind of (hardware or software) emulation will give such a drastic performance hit that you just cannot even think about it as an option.

if apple would switch to x86 hardware, or even began to support it, that would mean that all macintosh developers would have to re-code their software; which would effectively mean either abandoning macintosh or not supporting x86 hardware - so, you see, there's no point for apple in doing something like that.

Actually, I know quite a bit about technology, but thanks for the personal attack, its much appreciated. OS X is built on Unix, which runs on x86 hardware as well. I'm fully aware that the PPC platform is different than the x86 platform. I wasn't suggesting using emulation. A lot of program written on the Mac are also available on the PC.

Suck it up, you know the G4 is a pathetically slow processor. Apple should have switched to x86 gear a long, long time ago. The performance just isn't there, and quite frankly, the G5 isn't doing that much better.

DVW86
Sep 18, 2004, 02:57 PM
Suck it up, you know the G4 is a pathetically slow processor. Apple should have switched to x86 gear a long, long time ago. The performance just isn't there, and quite frankly, the G5 isn't doing that much better.

I use an 800MHz iBook G4 and a 900MHz Acer tablet PentiumM daily. Both computers have just over 512MB of RAM. I could not say that the Intel chip was faster than the G4. In fact I find that the G4 handles multi tasking much better than the Intel. Maybe its OS X, maybe it's the G4, but the iBook has far few problems and has a much smoother interface (less "stalls" and smoother graphics).

C-Mezak
Sep 18, 2004, 03:56 PM
I use an 800MHz iBook G4 and a 900MHz Acer tablet PentiumM daily. Both computers have just over 512MB of RAM. I could not say that the Intel chip was faster than the G4. In fact I find that the G4 handles multi tasking much better than the Intel. Maybe its OS X, maybe it's the G4, but the iBook has far few problems and has a much smoother interface (less "stalls" and smoother graphics).

You make a good point: while a mac might be slower under the hood, the experience at the user end is often faster than on a windows machine. The main reason for this is probably OSX. Im no ubergeek, and it's obvious to me after switching to mac that OSX is about a hundred times cleaner running than windows. UNIX was originally designed for machines that were meant to be rarely shut down, wasnt it? Anyway, I have friends with top-of-the-line windows boxes, who can barely get anything done now because their system has gradually become glutted and bloated. It may be possible for a geek to avoid this fate, but I think most computer users either have to reformat, or deal with a slow system on what should be a lightning fast machine.

Im sure that Apple understands that they have to keep the cpu upgrades coming, but there are plenty of other things they have going for them in terms of speed, IMHO.

charlie

dieselg4
Sep 19, 2004, 06:31 PM
Actually, I know quite a bit about technology, but thanks for the personal attack, its much appreciated. OS X is built on Unix, which runs on x86 hardware as well. I'm fully aware that the PPC platform is different than the x86 platform. I wasn't suggesting using emulation. A lot of program written on the Mac are also available on the PC.

Suck it up, you know the G4 is a pathetically slow processor. Apple should have switched to x86 gear a long, long time ago. The performance just isn't there, and quite frankly, the G5 isn't doing that much better.

Apple's current marketshare almost necessitates it to run on hardware OTHER than x86. If OS X ran on x86 hardware, that would effectively make apple a software company. It basically would only produce iPods, iSights, OS X, iLife, Final Cut, maybe a few other things. It's a bit of a catch 22 - OS X might be installed on many more computers if it ran on x86, but the revenue Apple makes would be cut drastically.

It wouldn't be all that different than OS/2 and OS/2 warp. OS/2 was, in many ways, a superior operating system to windows at the time. The development of OS/2 untilmately didn't pay for itself with sales, and it's now relegated to a blip in digital history.

slughead
Sep 19, 2004, 11:14 PM
you clearly know nothing or very little about technology; there's a reason why apple uses PowerPC processors and the consumer PC manufacturers use x86 processors - and that is called "binary compatibility".

you just cannot run PPC binaries on a x86 platform without any emulation, and i can assure you any kind of (hardware or software) emulation will give such a drastic performance hit that you just cannot even think about it as an option.

if apple would switch to x86 hardware, or even began to support it, that would mean that all macintosh developers would have to re-code their software; which would effectively mean either abandoning macintosh or not supporting x86 hardware - so, you see, there's no point for apple in doing something like that.

"waa you have to re-code your software"

hahaha, uhhuh, you're the expert, guy!


This is why I can take software designed for x86 linux OS's (like hx) and recompile it and run it on my G5, right? or maybe I didn't just do that 10 minutes ago.. yeah.. must've imagined it.

Or maybe this is how I can recompile things for x86 linux x windows and run it through x11

you must be right, instruction sets make alllll the difference heh.

I suppose you could be right though, binaries can't be emulated.. but ON THE SAME OS YOU CAN JUST RECOMPILE ON THE DIFFERENT COMPUTER.

Apple has already ported OS X to x86... or did you miss that Steve Jobs interview?

I'm not just talking about Darwin, either.. though Darwin for x86 is available for download for free.

If Apple released an x86 OS X on thursday, by friday afternoon 75% of all commercial software for OS X would be ported.

hah, recoding, that's a good one. I'm sure that extra 20 minutes of recompiling is going to really slow the wheels of progress.

Celeron
Sep 20, 2004, 10:10 AM
Apple's current marketshare almost necessitates it to run on hardware OTHER than x86. If OS X ran on x86 hardware, that would effectively make apple a software company. It basically would only produce iPods, iSights, OS X, iLife, Final Cut, maybe a few other things. It's a bit of a catch 22 - OS X might be installed on many more computers if it ran on x86, but the revenue Apple makes would be cut drastically.

It wouldn't be all that different than OS/2 and OS/2 warp. OS/2 was, in many ways, a superior operating system to windows at the time. The development of OS/2 untilmately didn't pay for itself with sales, and it's now relegated to a blip in digital history.

I'm not sure I follow your logic on this one? Apple doesn't make its current hardware, so why not switch to x86 and continue on as before? Instead of IBM floundering to make its G5, Apple could use AMDs Athlon64 line which is doing quite well. Instead of hobbling the Powerbooks/IBooks with the rather pathetic G4, they could just toss in a PentiumM.

I'm not saying Apple needs to stop designing the hardware, they just need to switch to the dominate, and faster, platform. There's no saying that Apple can't build the motherboards/bios in such a way that OS X will only run on machines equiped with such hardware.

And the day that OS X gets installed on x86 machines is when Apple truely starts making money. OS X is a pretty awesome OS, and its much better than Windows in many ways, but the short sighted view point that expanding their market share onto the x86 platform will ultimately hurt their bottom line is laughable.

dieselg4
Sep 20, 2004, 10:30 AM
And the day that OS X gets installed on x86 machines is when Apple truely starts making money. OS X is a pretty awesome OS, and its much better than Windows in many ways, but the short sighted view point that expanding their market share onto the x86 platform will ultimately hurt their bottom line is laughable.

Its not laughable at all. If they make signifacantly less money w/o the sale of hardware, they can't continue to develop OS X. I'm not saying its right, I'm just saying that currently one supports the other. Apple makes the bulk of its money selling hardware. That's a fact. The money, plus a smidge they make from final cut and other software, and the iTunes music store, and the massive amount they make from iPods, all gets redistributed to developing products and software.

I would like to see apple use faster hardware, but I remeber that when Apple licensed its OS out, it ran into the problem I have just described - Power Computing, Umax, etc., ate into Apple's hardware sales, which subsidized development of Mac OS, which helped put Apple into the red (further than it already was.) Basically, Apple doesn't sell enough copies of OS X to enough users for enough money to pay for itself.

Celeron
Sep 20, 2004, 12:22 PM
Its not laughable at all. If they make signifacantly less money w/o the sale of hardware, they can't continue to develop OS X. I'm not saying its right, I'm just saying that currently one supports the other. Apple makes the bulk of its money selling hardware. That's a fact. The money, plus a smidge they make from final cut and other software, and the iTunes music store, and the massive amount they make from iPods, all gets redistributed to developing products and software.

I would like to see apple use faster hardware, but I remeber that when Apple licensed its OS out, it ran into the problem I have just described - Power Computing, Umax, etc., ate into Apple's hardware sales, which subsidized development of Mac OS, which helped put Apple into the red (further than it already was.) Basically, Apple doesn't sell enough copies of OS X to enough users for enough money to pay for itself.

You aren't reading what I'm saying. Apple can keep making the hardware, keep developing the software with no licensing involved. Just switch from the PPC to the x86 platform. The machines out already use the same hardware for hard drives, opticals drives, memory and video cards. Apparently you aren't grasping what I'm saying. I'm not suggesting Apple drop their hardware line and simply make software for use on a Dell machine.

Silencio
Sep 20, 2004, 12:40 PM
"Relatively near future." Isn't that just more of the same Intel FUD we've come to know and love?

Multicore is one area in CPU design where IBM has a big advantage because they have been doing it for a lot longer than Intel, and multicore has been in the PowerPC roadmap all along.