To: Wolff who wrote (95915 ) 1/13/2000 2:32:00 PM From: Tenchusatsu Read Replies (1) | Respond to of 186894
Wolff, <The P7 is stillborn, the P8 will do well in servers> I don't think you've been listening, especially since you keep using the terms 'P7' and 'P8'. Here's a quick recap of Intel's roadmap: IA-64: Merced (mid-2000), McKinley (late-2001), Madison and Deerfield IA-32: Cascades (soon), Willamette/Foster (late 2000), other IA-32 proliferations. Intel's IA-32 roadmap will continue on for years. But Intel is simultaneously launching an IA-64 platform. The differences? IA-32 will obviously continue on the Pentium line where compatibility is king and the established software base is huge. But IA-64 will start a new line where compatibility is half-sacrificed and pure performance is king. <I will get out a AMD part pay less money buy off the self software and blow away the IA-64.........that what you imply, and remember in the corparate system arena legacy software can exits for years and years.> You still don't get it. IA-64 is not optimized for running IA-32 applications. No one is pretending that IA-64 will be a speed demon when running IA-32 apps. Software will need to be recompiled to take full advantage of IA-64. And as you imply, that's a very painful transition. But the transition is already taking place, with operating systems being run and debugged on IA-64 test platforms as we speak, not to mention high-end software packages. And you can bet that those native IA-64 apps will perform extremely well thanks to the new architecture of Merced. But if all you need is to run you existing IA-32 apps as fast as possible, well, that's why newer IA-32 processors like Willamette/Foster is being released. Let me explain it again. Merced (what you call P7) is NOT STILLBORN. In fact, it is very much alive and will be a vital spearhead for the new IA-64 line. But Intel isn't betting the entire farm on a brand new architecture like IA-64. Intel also realizes the value of compatibility, which is why Willamette/Foster (what you call P8) is necessary. It's a regular hedging of the bets, not an attempt to save face. The sooner you understand this, the sooner you'll be able to see what Intel has in store for the future. Tenchusatsu