Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Petz who wrote (59846)	5/27/1999 7:34:00 PM
From: Scumbria	Respond to of 1573901

John, In what timeframe will IA-32 be less than 50% of the market for PC's/workstations? Probably never. IA-64 doesn't offer any real performance advantage on the desktop, and any worthwhile implementation of IA-64 won't run x86 natively. Scumbria

To: Petz who wrote (59846)	5/27/1999 9:46:00 PM
From: dumbmoney	Respond to of 1573901

Should AMD be designing a CPU for the IA-64 intruction set or should they "roll their own" or stick with IA-32. What do you all think? That's an easy one. AMD should thank their lucky stars that Intel dropped the x86 ball, and continue x86 development. In what timeframe will IA-32 be less than 50% of the market for PC's/workstations? Quite possibly never. Not soon, in any case.

To: Petz who wrote (59846)	5/27/1999 9:55:00 PM
From: kapkan4u	Read Replies (1) \| Respond to of 1573901

<Petz - re: Should AMD be designing a CPU for the IA-64 intruction set or should they "roll their own" or stick with IA-32. What do you all think? In what timeframe will IA-32 be less than 50% of the market for PC's/workstations?> A very good question. I think they should stay with IA-32 for a long time (6 years or more), just extend the physical addressing of the instruction set to 48 or 64 bits and go for a very fast MHz design. I don't think they need to go beyond the 9-issue superscalar. Kap.

To: Petz who wrote (59846)	5/27/1999 9:58:00 PM
From: grok	Read Replies (1) \| Respond to of 1573901

Re: <In what timeframe will IA-32 be less than 50% of the market for PC's/workstations?> Micro Design Resources, the folks who do Microprocessor Report and Microprocessor Forum, have a projection that shows that IA-64 first achieves about 10% of the total IA-32 market in about the middle of 2003 and then achieves about 100% by the end of 2004. 50% would occur sometime in 2004 according to them.

To: Petz who wrote (59846)	5/27/1999 10:22:00 PM
From: kash johal	Respond to of 1573901

Petz, Re: "Should AMD be designing a CPU for the IA-64 intruction set or should they "roll their own" or stick with IA-32. What do you all think? In what timeframe will IA-32 be less than 50% of the market for PC's/workstations?" I think that if you look at whats happening there will only be 2 viable 64 bit architectures - Alpha and IA 64. The infrastructure for a third seperate architecture is impossible for AMD at it's current size to even contemplate. SO my best guess is that K8 will be essentially be an Alpha with a hardwired x86 engine for 32 bit apps and the two CPUs on die would share an onchip bus similar to the EV6 bus. This would have an advantage for server customers as the Alpha platform has a huge infrastructure from Linux to NT to chipsets to proven code. If you take the die size of Alpha in 0.18 say 150mm2 and die size of K7 say 100mm2 and simply add them you get 250mm2. current estimates of merced are in 200-300mm2 range for 0.18 process. So in a 0.13 process this monster may run 160-180mm2- or current K7 sizes. I would expect a 0.13 process from AMD possibly by early 2001. Just my best guess right now. I think the K8 will be presented in October at their annual microprocessor shindig and we should know more then. Regards, Kash

To: Petz who wrote (59846)	5/27/1999 10:26:00 PM
From: grok	Read Replies (1) \| Respond to of 1573901

Re: <Should AMD be designing a CPU for the IA-64 intruction set or should they "roll their own" or stick with IA-32.> Well, they certainly should not "roll their own" since this would take a vast amount of resources and time, just like it's taking for Intel on IA-64. To design for IA-64 may be impossible since you can be sure that Intel and HP have been so careful with patents this time that it would be very hard for anyone else to provide an IA-64 chip and even harder for anyone to use it in a system. So that leaves stick with IA-32 as the only option. Since it is likely that Merced and its follow-ons will still require years before they achieve much market penetration, sticking with IA-32 will be fine during this period. Of course, if and when the day comes that IA-64 becomes dominate it will be curtains for AMD.