• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

How well do you think this would fold?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
:drool:

Hmm, cancer solved in 10 years? ;)

But wow...if only i could get my hand on one of those......

/me thinks evily ;)

Brian
 
Ok, so if Team 32 could pull a heist on this thing, you think we'd be able to keep it a secret once we started cranking out WU's in fractions of seconds? ;)

Think of it, the bottleneck would be the connection to Stanford, lol.
 
Anteraan said:
Ok, so if Team 32 could pull a heist on this thing, you think we'd be able to keep it a secret once we started cranking out WU's in fractions of seconds? ;)

Think of it, the bottleneck would be the connection to Stanford, lol.
We'd have to run over 10,000 instances, could that even be done? Or we could run like 2 instances per CPU so like what, 22k? That's just nuts.
 
Does the Hypertransport feature allow the CPUs to be clustered into virtual processors? If so, the 4 processor limit kicks in, so we'd have to group 2500 CPUs into each processor.

Let's see ... 2500 CPUs at 2GHz each ... I don't think even the largest Gro WU would bring it to 100%. It'd be spitting out completed WUs like peanut shells. :D

So ... this thing's ... where ... at NASA? Ah! Sandia National Labs ...

Anyone know anybody there? :D I mean, they can't simulate ... secret stuff ... 24/7 ... can they?
 
"Opteron cluster at Sandia National Labs," << Loud =)
i used to live in the headquarters of Phillips Petroleum and one of their IT guys was going to show me through their computer place... they had Cray and Beauwulf clusters all over the place however he had to move to China unexpectedly for work and so i never got to see that wonderful place =)
 
(1) Since Vijay keeps reminding us over at the support forums that FAH Is not SMP enabled, and since I have an idea what Sandia will be doing with this monster, I think we'd have to load 10K or 20K FAH images. Suppose we could figure out how to manage it remotely?

(2) Why are they using SDRAM? :eek: I've heard time and time again that Athlon CPUs really like DDR.
 
Back