[yoyo52]

I've heard that IBM currently has G4 processors that run at 1.6 GHz, but without the velocity engine, which is controlled and licensed (or not, which is actually the case) by Moto.

Some questions about this.

1) Is that rumor true?
2) If it is true, would you accept a non-velocity engine IBM processor running at 1.6 GHz, or would you absolutely demand to have a velocity engine enhanced processor?
3) Do you think that Apple should allow people to make the choice of which processor they want--have BTO options for either Moto processors with velocity engine and IBM processors without?

Finally, a question about Moto: why don't they get off their butts? I mean that seriously--is there a cogent reason/reasons why they can't or won't provide faster processors?

[nonhuman]

Originally posted by yoyo52:
<STRONG>I've heard that IBM currently has G4 processors that run at 1.6 GHz, but without the velocity engine, which is controlled and licensed (or not, which is actually the case) by Moto.

Some questions about this.

1) Is that rumor true?
2) If it is true, would you accept a non-velocity engine IBM processor running at 1.6 GHz, or would you absolutely demand to have a velocity engine enhanced processor?
3) Do you think that Apple should allow people to make the choice of which processor they want--have BTO options for either Moto processors with velocity engine and IBM processors without?

Finally, a question about Moto: why don't they get off their butts? I mean that seriously--is there a cogent reason/reasons why they can't or won't provide faster processors?</STRONG>

1. I sure hope so.
2. I would definitely go with a non-velocity 1.6 GHz G4 if the GHz rating is consistent with current processors and not just an inflated number based on longer pipelines or something like that. If a 1.6 GHz G4 is actually twice as fast as an 800 MHz G4 then who really needs a velocity engine. I don't have one now in my 500 G3...
3. YES! Apple definitely needs to provide more options.

[THT]

1. The rumor is false. Or perhaps you're confusing it with the Power4 which wouldn't ship until 2002 at a minimum. It could be 2003 for the Power4 considering how fast IBM moves. The best tidbit about an IBM desktop class PowerPC is from this eet.com story:

IBM mulls memory changes to PowerPC

... For its next-generation architecture, IBM is working on a single-chip multiprocessor PowerPC which the company describes as a multicore superscalar design and which will work in symmetric multiprocessor systems.

The multicore processor could run "a single instruction stream or separate instruction schemes", said Gala.

The 1GHz-plus multicore superscalar processor will incorporate a single-instruction, multiple-data engine along the same lines as Motorola's Altivec and support for RapidIO.

"We are talking to customers about what they would like on top of Altivec," said Gala.

Now, they could be talking about the Power4 here which would be a very expensive chip or it could be a desktop class PowerPC. Who knows. What we do know is that neither shipping. The interesting thing is what would a server chip need an SIMD unit for? So, the info above seems to point to actual CMP PowerPC for the embedded and desktop markets.

2. There's no reason for IBM to exclude an SIMD unit in their future embedded or desktop PowerPC processors. There's is not one good reason whatsoever not to have an SIMD unit in that market. The embedded market demands it. Great DSP performance is what the embedded market is about, and the performace gain with SIMD units (like AltiVec) is very good. Desktops aren't too far away from demanding it either.

3. No.

Motorola's PowerPC market is the embedded market. Apparently, they are getting the right amount of business there not to warrant going full out on a PowerPC design.

[ironknee]

Why is Moto making deals with Cisco supplying chips when they can't keep up Mac demands?

damn it

[Scott_H]

Originally posted by ironknee:
<STRONG>Why is Moto making deals with Cisco supplying chips when they can't keep up Mac demands?

damn it </STRONG>

'cause they want to make money?

[Evangellydonut]

Originally posted by yoyo52:
<STRONG>I've heard that IBM currently has G4 processors that run at 1.6 GHz, but without the velocity engine, which is controlled and licensed (or not, which is actually the case) by Moto.

Some questions about this.

1) Is that rumor true?
2) If it is true, would you accept a non-velocity engine IBM processor running at 1.6 GHz, or would you absolutely demand to have a velocity engine enhanced processor?
3) Do you think that Apple should allow people to make the choice of which processor they want--have BTO options for either Moto processors with velocity engine and IBM processors without?

Finally, a question about Moto: why don't they get off their butts? I mean that seriously--is there a cogent reason/reasons why they can't or won't provide faster processors?</STRONG>

G4 w/o velocity engine is another G3. G3 with longer pipe to push up speed is just like P4's way of doing things...you can argue that 7 stages is not nearly as bas as 20, but 4-&gt;7 is a 75% increase, not much less than the 100% increase of the P4 over P3 (and only 54% increase over Athlon). IBM's G4 was entirely different from Moto's, and even if they sold those parts to Apple right now, you won't see it in machines for another year or so since new Mobo has to be designed 'n what not. so
1. could be true, but makes no difference.
2. in RAW power, 1.6 Ghz in the lab is still slower than 1.8 Ghz P4 on the market, and behind by a few months in release date. And when Intel and AMD are both putting SSE/SSE2/3DNow (pretty much a modified subset of SSE) on their chip to handle Photoshop (among other things), not having it on the G4 can be potentially very harmful for the graphics market.
3. Do you want a slower yet development cycle due to halving the Apple HW team in half to support 2 different CPUs on different mobo? Do you want extra design cost that results from it?

It took IBM quite a while to push their 4 staged chip to 700Mhz as you see in the iMacs today (and probably upto 850 or more in the labs), while Motorola re-designed a chip, and was able to mass-produce it. I think the biggest reason we are so much behind now is that Moto spent too much time trying to push the 7400 chip to more than 500Mhz rather than give up on it early, realizing their fab is not as good as IBM, and start designing the 7450. The time it took between PIII 1.13Ghz recall to 1.5 Ghz P4 hitting the market was considerably sorter than G4 500Mhz downgrade to 450 and G4+ hit the market at 667, (less than 6 months for Intel and more than a year for Motorola) that is where the problem lies...I'm not advocating that Moto should make a half-baked CPU like the P4, just saying that their mistake took place back then so figure out where to point fingers next time.