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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: dale_laroy who wrote (59493)	10/20/2001 10:53:09 PM
From: Tony Viola	Respond to of 275872

Dale, nice, objective post. Your 3X factor is even generous, IMO. Historically, a processor has had to deliver three times the performance of an entrenched architecture to dislodge it. For all intents and purposes, Itanium can be considered and entrenched architecture. As for Itanium's higher costs, while this is true of Intel's traditional Xeon market sectors, Itanium promises a significantly lower cost than was attained by the MIPS, PA-RISC, and Alpha systems it will be displacing. For x86-64 to draw the market away from Itanium in these high end markets, x86-64 will have to deliver at least three times the performance of Itanium. Tony

To: dale_laroy who wrote (59493)	10/21/2001 12:32:14 AM
From: pgerassi	Read Replies (3) \| Respond to of 275872

Dear Dale: You might be right, if the Itanium was established. It is not. It needs about another 5 years of leading performance to do. It still has competition from PA-RISC, Power4, SPARC, Alpha and even its sibling, Xeon. If it does not leap in front of the pack by the end of next year, it is history! The rats will leave the sinking ship and it will be quickly forgotten except by those who lost money, big time. The entrenched architecture of x86 can be quickly overcome, if it is compatible. Note how quickly the 386 was accepted from the 286s it replaced. Or the Z80 from the 8080. Incompatible replacements need 3 times, but compatible ones can do it with as little as 50% more so long as the old mode is at least as fast. If the new and old modes are faster on the new one than the old one, an increase of 25% is all that is needed. If the performance is the same, but the price is 1/3 and the software already exists for the alternate platform, the changeover is quick (eating from bottom to the top has been done, just look at the 386 and later CPUs). I just do not think that Itanium is established to the degree you think. There is a million PA-RISCs, a million PowerPCs, a million Alphas, a million SPARCs and many million x86 servers. There is not a million Itaniums and not likely to be for 5 years even in the most optimistic scenario. So Itanium is not tried and true. Heck it has yet to pass being accepted by Corporations as a whole (it has not been 12 to 24 months since release in quantity) and even after that it takes 3 more years for enough history to become established. Heck, Itanium has yet to match the performance in normal server applications. Thus, it doesn't make the three times performance you say it needs to before becoming established. Rambus had the same advantages you say, but it has not yet been established. In fact, more and more run away from it. Itanium will be no different, especially if it doesn't live up to the promises, much less the hype. If x86-64 draws half of the x86 market away, it will be far more established than Itanium would ever get. And the 64 bit mode will have far more hardware available to run it than all other 64 bit GP CPUs combined. The pitch is easy to see: "Look, this CPU runs all of your 32 bit x86 software, as is, faster than you do now! No retraining, no updates and no additional software investment required! And for less than 5% more, you get 64 bit capabilities enabled and even more performance, while still doing your old applications with no loss of speed!" And that gets customers on your side by making them want to do it rather than having to shove it down their throats. The later course almost always fails sooner or later. Pete