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.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: tcmay who wrote (164701)	4/30/2002 1:02:15 AM
From: Jim McMannis	Read Replies (1) \| Respond to of 186894

RE:"Intel bit the bullet and went for the longer pipeline, thus initially being a bit disappointing when the speeds were around 1.6-1.8 GHz. But now that the speeds are coming out at 2.4 GHz and higher, even 3 GHz, air-cooled, the "breathing room" is turning out to be as promised" Yes, exactly. And while some of us were trashing the IPC of the P4 Intel improved that with Northwood (512 cache) and made up the rest of it with Mhz. On top of that. Mhz is the single most important marketing factor to the masses. Yes, AMD should have got a clue with P4s longer pipeline...even giving them the benefit of the doubt they should have tried to use a PR rating long before they did. Some here laugh at the PR rating but I see no other choice for AMD. RE:"Likewise for the IA-64. AMD may well get some desktop sales in 2003-4 for their Hammers, but only as a "stopgap." The x86 core architecture cannot plausibly be extended for the next 10 years and extended for use as a building block for high-end machines. (Note that none of the current "mainframe" computers are based on x86: IBM has the Power4, Sun the SuperSPARC variants, HP _had_ PA-RISC and was working on VLIW before teaming with Intel, and Compaq/DEC had the Alpha. None of them were planning to use x86 architecture for enterprise/transaction processing computers.)" This is a very good point and has crossed my mind more than once. It does appear that x86-64 will take AMD only so far. Maybe an easy transition but perhaps an eventual dead end. I have my doubts that AMD has the gumption or resources to move to a new architecture. Seems they leave that up to Intel and then follow. Even when they come up with something good like hypertransport they still seem to back off making it a standard in favor of Intels standard. RE:"I predict that AMD will realize, soon, that it needs to move to longer pipelines in its Athlon line. And it will realize in about 3 years that the Hammer is a dead-end." I don't know enough about chip design to know if they can even do that (increase the number of pipeline stages)and if so, why haven't they already done it. If the hammer is dead end in a few years isn't the Athlon dead end as well? Or any x86 architecture for that matter? What is the architecture for the masses going foward from the limitations of x86? I did appreciate your expose on the poisson distribution for approximating chip yield. I was using a somewhat more primative estimator for the number of good die per wafer so I will play with your formula. The HP calculator it itself was quite a luxury back in the mid seventies. At UCSB they stuck us in a room with a bunch Wang calculators with what appeared to be neon displays. Those who could afford an HP-45 had a definite advantage. Jim