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: jspeed who wrote (210687)	9/12/2006 5:02:25 AM
From: Rink	Read Replies (2) \| Respond to of 275872

Jspeed, re: Rev G will definitely be better clock-for-clock than rev F. Per AMD's analyst day presentation, 50% better floating point performance. Int performance is improved too, but it's not clear how much. What slides to you mean? I looked it up once again and I see nothing that confirms that rev G has the additional FPU. Here are a couple of examples: Apparent rev G waver without additional FP unit (this was confirmed by Hans de Vries iirc): epscontest.com Next generation die sketch with added FPU, added HTT channel, added L3, added power mgmt, etc... (K8L, not rev G, fafaik that is): epscontest.com The situation isn't clear however. Above 'next generation' may be introduced in phases. Possible phases include: 1- QC based on rev G 2- QC based on K8L on the rev G sockets and platform (no use of 4th HTT channel, no use of FBDIMM, possibly even no use of HTT v3, but with the added FPU, and with the added L3). 3- QC based on K8L on new K8L platform. AMD has been sufficiently evasive to let investors guess. That is by the way exactly what a recent discussion between Doug, Mas, FPG on this subject showed. I think Doug made the point that AMD could have made the claim that "the 1st QC will have a 128b FPU and is due summer '07" explicit, and because they didn't they might still choose to go QC without added FPU first, followed by the real McCoy. Without additional data it is impossible to make a fail safe conclusion. For reference I just looked up where that discussion started and I think it was here: Message 22792557 Anyways, based on how I interpreted the data I think first rev G is basically a rev F shrink and will not contain an additional FPU. It won't be much higher performing either, but will allow to shift to DC almost completely. Besides that I think the first QC might be K8L (point 2 above) but it is still quite possible that it is a QC rev G instead (point 1). Regards, Rink PS: Maybe you made a typo or something as here is where you claim that K8L (not rev G) has the added FP unit: Message 22795095