IBM claims that Apple's statement about Intel being more powerful isn't true.

[Convert]

http://www.eweek.com/article2/0,1895,1833126,00.asp

VERY interesting. What does everyone think? I think Apple left IBM for more reasons than power...

 

[CreativeEye]

all well and good after the fact!

of course they'll say this stuff now after not delivering to apple and have bad stock supplies.

'...IBM PowerPC chips could cover Apple's entire product line, Adkins said...'

short term we know the answer is 'no they can't'.

long term we don't yet know what chips intel will have for apple...

 

[Convert]

Exactly. That was the main problem with IBM, they simply couldn't deliver what Apple wanted (based on speed as well as simple request). Not to say they're a bad company, just not good enough for the needs of Apple.

 

[ElDiabloConCaca]

Right -- what good is a company that says, "Sure, in theory, we could blow them away, but in actuality, we don't currently manufacture a processor that does."

Look here:

But IBM begs to differ. The company could build PowerPC chips that satisfy the needs of the entire range of Apple's product lines, including portables such as the PowerBook, said Rod Adkins, vice president of development for IBM's Systems and Technology Group, which produces IBM's PowerPC chips.

Keyword: could.

That means that the chips aren't here yet.

I have no doubt that IBM really could delivery chips for every product offered by Apple, but the problem is that promises don't go far anymore -- IBM suggsted to Apple that they could have 3.0GHz by a certain time and have yet to deliver, years later.

If you can make processors that will be a better choice than Intel's, and you want Apple's business, then build the processors. Saying you can isn't the same as doing it.

 

[Mikuro]

Their desktop G5s still beat the P4, and they just released a low-power G5. Looks to me like they can and have done it (although granted, the low-power G5 doesn't run at very high clock speeds, so maybe it wouldn't really tip the scales anyway).

Personally, I've thought from day 1 that any hints at IBM's poor performance were just spin to help the public swallow the announcement. Jobs didn't even really say it was about performance, he just kindasorta implied it. He had to say SOMETHING, and I think the real reasons for the switch are too far off to be spilling the beans on yet. So he made IBM out to be a lot worse than they really are.

If I were IBM, I'd be upset, too.

 

[Veljo]

I agree EXACTLY with ElDiabloConCaca in this:

Right -- what good is a company that says, "Sure, in theory, we could blow them away, but in actuality, we don't currently manufacture a processor that does."

They promised a faster processor a long time ago and even now have still yet to deliver.

Intel are clearly the leaders in processor technology, and because their sole focus is with processors they are the ones who are more likely to excel. Either way I know how IBM feel, but they just did not deliver and that's essentially what lost them their position in Apple products.

 

[nixgeek]

I don't know about leaders (especially since AMD is currently mopping the floor with Intel thanks to AMD64 on x86-64 performance). I think it's more related to Intel being able to meet the demand for Apple's chip needs. How many times have IBM and Motorola held Apple back due to their shortages in PPC chips? Intel, while probably not the fastest CPU brand, is by far the most available especially when compared to AMD, and right now their CPUs are much cheaper than some of the Athlon 64 CPUs out there.

In short, Intel has the money and fabs to give Apple what it wants. Intel also needs to hedge its bets with Apple especially with what they are experiencing with Microsoft's choice for AMD64 over Intel's EM64T. This is their opportunity to shine with Apple and vice-versa.

 

[fryke]

It's also a bit difficult. With Apple actually being the _sole_ computer vendor to actually make use of the PowerPC as a desktop and notebook processor (yes, there are others, but they are _very_ small players and for this thought don't even matter), we only see what IBM can deliver if Apple shows us. IBM, for what it's worth, could deliver a G5 notebook processor running at 2.5+ GHz right _now_, and as long as Apple doesn't produce a computer that makes use of it, for what we know, IBM didn't deliver.
Once Apple jumps boat, IBM simply _won't_ deliver desktop and notebook PowerPC processors. And we won't know where they would have driven. We'll only see what they'll do with the POWER series and with the console chips (and those are updated every other year at best).

Btw.: I thought we already discussed this about two weeks ago or so...

 

[scruffy]

There's a rather credible article on ars technica
http://arstechnica.com/columns/mac/mac-20050710.ars
That puts forth a possibility that no one here seems to consider - IBM told Apple they weren't interested in their business.

Lots of people here would likely get offended at the mere suggestion, but they do make a credible argument. Worth reading anyway.

 

[HateEternal]

Their desktop G5s still beat the P4, and they just released a low-power G5.

Exactly. The majority of Benchmarking that I have seen between x86 and PPC for say photoshop performance and such always pit the G5 versus a x86 server class chip. It is always G5 vs Xeon or G5 vs Opteron. They don't typically compare the G5 to a desktop class chip like the P4 or an affordable AMD line.

And yes, the Xeons and the Opterons are faster, but I don't think we are going to be seeing PowerMacs coming out with Intel's server class chips.

 

[fryke]

*sigh* - and people stay comparing _today's_ G5s and _today's_ Pentiums, Xeons and AMD x64 chips. Yet clearly, Apple's not going to use those in next years PCs. Well, maybe the G5s, because IBM still doesn't deliver? But intel quite surely will deliver other parts. intel's roadmap is out there. Read it, it's interesting.

 

[HateEternal]

*sigh* - and people stay comparing _today's_ G5s and _today's_ Pentiums, Xeons and AMD x64 chips. Yet clearly, Apple's not going to use those in next years PCs. Well, maybe the G5s, because IBM still doesn't deliver? But intel quite surely will deliver other parts. intel's roadmap is out there. Read it, it's interesting.

It doesn't matter if its today or tomorrow. Today the PowerMac is comparable to a x86 server class machine, when the x86 PowerMacs come out are they going to compare to desktop class PCs or Servers?

 

[mindbend]

I imagine with a decision as huge as this, it was a variety of factors.

1. Current speed
2. Roadmap speed
3. Pricing structure
4. Incorporation of support technologies (e.g. video cards, controller chips, etc.)
5. Which side of the bed Steve rolled out of that day
6. Heat requirements
7. Flexibility

The list could go on and on. I doubt the decision was 100% speed-based, but I imagine it was a big part.

I think the main reason for the switch is very simple. Apple can piggyback the more widespread technology that Intel will be developing. There will be cost benefits to this (cheaper) and cutting edge benefits as well as implementation benefits (faster to develop).

 

[fryke]

It doesn't matter if its today or tomorrow. Today the PowerMac is comparable to a x86 server class machine, when the x86 PowerMacs come out are they going to compare to desktop class PCs or Servers?

Of _course_ it does matter. If you listen to what Steve said in that WWDC keynote, then his reason was "power per watt", and that intel had a compelling roadmap, whereas IBM did not. So Apple did not choose today's intel processors. Neither server or desktop or notebook. But the desktop processors that will come out sometime next year.
On the other hand you have IBM. They took the POWER series, cut it in half and built a "desktop" processor with it (the original G5). The original 2 GHz processor needed quite a bit of inventive cooling technology. And when the time came when we (plus Steve Jobs) expected 3 GHz, we got a 2.5 GHz 970FX that needed active liquid cooling. This doesn't look "cool" to me (performance/watt). Well: Neither do Pentium 4 processors, I agree, but those are not what Steve is looking for at intel.
The timing _does_ matter indeed.