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: peter_luc who wrote (83044)	6/20/2002 12:00:57 PM
From: wanna_bmw	Read Replies (3) \| Respond to of 275872

Peter, I'm trying to put a realistic view of Hammer for myself. Regarding your points. Re: "Sledgehammer will bring high performance computing to a price range never seen before." Yes it will, but how much of a difference will the price of the CPU make towards the whole system. Itanium 2 servers will apparently start around $40k for a 4-way system. These processors will probably be in the low end of Intel's enterprise pricing, or about $1200. (I am using Intel's current prices for Itanium as a guide intel.com. If Sledgehammer sells for half of that, then they cumulatively save $2400 on 4 processors, and maybe another $400 on the cost of the platform. Add 20% to that to make up for the difference in margin on the server and you get a cost savings of $3360. Seems like a good savings, but that still puts the price of the 4-way server at $36.6k, so the overall savings are about 8.5% (and this is with Sledgehammer at 1/2 the price of the lowest grade Itanium 2). In terms of performance, Sledgehammer may beat a low grade Itanium 2 (900MHz with 1.5MB L3 cache), but I think it will depend on the application. With performance metrics on high end software a complete mystery, I don't think we could speculate on Hammer vs Itanium 2 performance. However, Itanium 2 will have 64-bit software ready for it, while similar software for Hammer will take a while. I doubt that current RISC customers will be willing to transition their current 64-bit apps over to 32-bit x86 in anticipation for an x86-64 revolution in the future. That's just too many transitions for them. Itanium 2 only means one transition for them. For other x86 customers, the choice may be easier, since they won't have to transition at all for Hammer, except for when 64-bit apps arrive in the future. So the decision will be whether to transition to 64-bit later for Hammer, or 64-bit now with Itanium 2. The latter will be more expensive, but Intel seems to be guaranteeing a long lifetime for the design, so that may justify the expense for some. For others, they may choose Hammer, or they may stick with Xeon. Re: "Clawhammer will even bring 64-bit computing to the masses. This may turn out to be a compelling concept." Yes, the concept is compelling, but there is a problem. I haven't heard anything about a release date for Microsoft's Windows 64 for x86-64, let alone any software announcements from developers on x86-64 ports of their applications. Therefore, I can only assume that this will be late in coming. I think that high end applications are more likely to get ported before consumer applications, but that's just a guess. Re: "Let's hope that the Hammer will not slip to 2003. Time to market may be crucial in this situation." I agree. TTM is a main concern for AMD with Hammer. They need to deliver it, but deliver it robustly, too. I don't agree with an early Hammer launch unless they can guarantee flawless execution. wbmw