SI
SI
discoversearch

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)
AMD 210.80-4.8%3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Paul Engel who wrote (17607)11/4/2000 4:31:30 PM
From: fyodor_Read Replies (2) of 275872
 
Paul, serious question for you:

What do you think the Itanium will be (s) best suited for? I do a lot of double precision fp-intensive (quantum) calculations (on huge datasets) and was wondering whether the Itanium (or McKinley) would be suited for that.

I would love to be able to use MP Xeon boxes, but the problem is that the P6 core just isn't very good at double precision (it runs the calculations in question roughly half as fast as a K7). I was going to build a cluster of Athlons, but I've been really bogged down with other matters and haven't gotten around to it yet. I'm also considering the P4. There's plenty of ILP (and thread-level same) in the calculations, so SSE2 might be the way to go... (I don't want to rewrite the code for only a small boost, so I won't commit to the P4 before seeing some relevant benchmarks).

Depending on P4 system price, Athlon clusters might still be the way to go, though... (we're seriously feeling the squeeze around here, having suffered a 25% budget decrease over the past couple of years).

TIA,

-fyo
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext