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: Joe NYC who wrote (78463)	4/26/2002 4:39:45 PM
From: Petz	Read Replies (1) \| Respond to of 275872

Pricewatch now has PC3200 DDR on the front page 256M=$86, 512M=$195. <<Dual channels would also require some incrase in latency.>> <Why is that?> For the same reason that CAS 2 usually won't work with 2 DIMM's instead of 1. Imagine with 4, 6 or 8 DIMM's installed. < Now, with all the benchmarks rigged this way, and AMD fully aware of the fact how the benchmarks have been rigged, AMD knows where the trap is, but they are walking right into this trap.> Would it have been wise to count on Sledgehammer beating it's schedule, like it apparently did? Even having working silicon now, it would take a lot longer to qualify motherboard designs with 2 64-bit channels compared to one. It's pretty obvious to me that if Clawhammer arrives in October, it won't take 6 more months -- more like 3 -- to get to Sledgehammer. What would it give AMD to drop Clawhammer? If they could speed up SLedgehammer intro by one month, they'd be lucky, so they'd still be hammer-less for two months. EDIT - And with the 0.13u process apparently requiring several iterations to get good yields, it behooves AMD to start with the smaller-die-size Clawhammer before the larger Sledgehammer. Don't forget that dual core Clawhammer systems WILL have two channels of 64 bits. Finally, with PC3200 already available, I have no doubt that Via and SiS will support PC3200 for Clawhammer before 2003. Petz

To: Joe NYC who wrote (78463)	4/26/2002 8:21:02 PM
From: pgerassi	Read Replies (1) \| Respond to of 275872

Dear Joe: Did you look at the slides in the meeting presentation? Clawhammer SCSC gets 135% of the memory bandwidth of AXP at the same clock. Thus, since AXP gets 2.1GB/sec, Clawhammer gets 2.84GB/sec with PC2100 as base effective. With PC3200 (you know this is available now), Clawhammer would get an effective 4.46GB/sec faster than NW@133/533 dual DDR PC2100 or 4.2GB/sec. Clawhammer thus will have higher effective bandwidth than NW. Opteron (DCSC) gets 205% of AXP or 52% more than Clawhammer. It would get an effective rate of 4.3GB/sec with PC2100 and 6.56GB/sec with PC3200. Clawhammer (suggested name Athlon XP64) will have enough effective BW to outrun NW. Yes, I know that Opteron (Hammer DCSC) trumps NW in spades even in 1 way workstations. Pete PS: In a previous post Itanium 2 is said that the memory expansion controller will double bandwidth to 12.8GB/sec. That means that normal bandwidth is 6.4GB/sec for Itanium 2 systems. Single Opteron DCSC will have a higher effective BW than Itanium. 2 way Opteron will outrun the maximum of an Itanium 2 SMP system. More than 2 way Opteron systems would pulverize Itanium 2 systems in BW limited tests like SPEC. When DDR2 becomes widely available with speeds on the roadmap to 800MHz, even Clawhammer would have more bandwidth than Itanium 2, All Opteron systems of any number of ways including single way would beat Itanium 2 systems.