TR Forums

Well i've devoted a core to my two 3850s now and the results are pretty impressive


on the three remaining cores [at 3.1ghz] you get about a WU every 20 hours. Not too shabby, there's better performance out of C2Qs but what makes up for it is the crossfire'd videocards pulling their weight
GPU work Units go fast.
VERY fast.
I get a percent every 2 minutes.
In an two hour work unit worth between 100 and 200 points [well 98 and 196] that adds between 1200 and 2452 ppd to my total
that on top of the roughly 1900 ppd i get out of the quad already, you can build a nice machine with a radeon or two.

Hmm... Assuming you can get a max of 4000ppd out of a C2Q, then if GPU adds 1200ppd then you will be at 3100ppd. If it brings 2452ppd then you will be at 4352ppd. So depending on the WUs you can either be better or worse off. I am sensing that after the points reassignment we will have balance.

Now the question becomes whether the increased ppd is worth the power draw. That we will require more data.

Flying Fox wrote:

Hmm... Assuming you can get a max of 4000ppd out of a C2Q

Yea, VMWare player has been choking and dying so this is with the windows smp client. Im also benching from 2665 WUs to be conservative

Aye, I havn't looked into the economics of the power draw as my cost efficiancy is "I can have one computer drawing as much as I want and one laptop, and I can't put any more computers in"