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: kash johal who wrote (52976)	8/30/2001 11:11:32 AM
From: fyodor_	Read Replies (4) \| Respond to of 275872

Kash: I think the multi-threading stuff is great. I believe it already exists on SUN CPU's for a while as well as Alpha. So there is already SW that utlizes such capability. I believe the Alpha chip that will support SMT is the EV8, i.e. not one shipping now. One of the advantages of SMT is that it can be made in such a fashion that any software that can utilize multiple processors can use it. Instead of having two separate physical processors, you just have two virtual processors. At 0.13 PIV will be a real winner. [...] AMD is toast IMHO. Gotta love the extreme optimism / extreme sceptisism duality of this thread ;-) I think you are overstating the significance, as well as underestimating what AMD has on the drawing board (Sledge). Remember: The consumer version of WinXP won't support multiple processors (and therefore won't support Jackson Technology). This goes well with earlier indications from Intel that only the server version of P4 would use SMT. Additionally, virtually zero consumer apps are multithreaded, so even if XP was to support it, apps wouldn't be able to make use of it. As far as AMD goes, they have several patents in the area and the "dual core" / "not dual core" of the Sledgehammer could well turn out to be SMT. AMD is currently around 30% ahead in IPC (1.4GHz Athlon ~ 1.8GHz P4), but behind in operating frequency by 40%. Assuming AMD has gone the route of extending the pipeline in Hammer, a 40% increase in frequency could likely be attained without sacrificing all of the IPC lead (however, I don't think they will attain a 40% increase, but that's another matter). However, AMD has a lot of low-hanging fruit as-yet unpicked in the Athlon. Only a small portion of that is picked with the Palomino. E.g. the L2 cache still sucks, both in terms of latency and in terms of bandwidth. This business is complex and, at least in terms of outlook, very cyclical. Periods of clear outlook (although not necessarily good), where the processor / architecture / platform is known for at least a few years into the future, give way to periods of cloudy outlook, where investors remain mostly in the dark about the new products on the horizon, fuelling much speculation. -fyo