SMPand Multi GPU's coming soon

zim01

Gawd
Joined
Jul 24, 2006
Messages
1,004
"November 2006 The SMP client is now looking good enough that we are starting a more broad alpha test outside of Stanford. If that looks good, we will move to a broader beta test of this new client"


"10/27/2006 Multi-gpu support coming soon
We are testing a new client/core (core v0.09) which allows one to use multiple GPUs in one box. We hope to release this next week".
 
I can't wait for this so I'll put my X2 to a better use :) FTW

Looks promising :)
 
i haven't heard much about this.. will it offer more performance or just more convenience? either way its a good thing, I'm just wondering.
 
It'll likely be able to handle more complex/longer calculations/simulations in a reasonable timeframe as it will run as a multi-threaded core. Would be cool to see how it takes advantage of that 8 core monster one of the guys around here has running! :D
 
Nice.. I can't wait for that.. I wonder how many years it will be before they introduce true distributed computing for home or office networks.. It would be an interesting way to perform larger, more complex tasks for the horde.
 
Woohoo! And maybe the multiGPU client will fix the CPU usage issue. I've got a quad box that'll be on when I figure out how to keep it quiet enough - looks like it'll be a test box for multifolding too :D

 
i wonder if there would be any advantage of running the SMP vs two of the regular consoles... I already have almost 400 dual core machines each running two of the regular consoles so it would kind of suck to have to go back and do it again :(
 
FLECOM said:
i wonder if there would be any advantage of running the SMP vs two of the regular consoles... I already have almost 400 dual core machines each running two of the regular consoles so it would kind of suck to have to go back and do it again :(

If they're all on the same network then just script it. Back whoopin' ass before ya know it ;)
 
They're saying client, but make no mention of SMP cores. I wonder if that means the client will autodetect and setup two instances and watch both, or if the cores too are SMP aware.

 
mess on dual core boxes with GPU folding too. Need to leave 1x (or x% of the 1st core open) open for the GPU how's that gonna work?

Always good moving forward.
 
I've a box that Dell replaced - It was a Dual Xeon, they replaced it with a dual Xeon, dual core, with HT - went from 2 CPU's to 4 plus 4 virtual. Would like to try the new SMP on it:)
 
Back
Top