It’s getting cold, time to fire up the GTX1080

jebo_4jc

[H]ard|DCer of the Month - April 2011
Joined
Apr 8, 2005
Messages
14,566
Hey guys. Been a while.

I loaded up the Windows FAH client and I’m getting around 500-600k PPD. Is that an optimized production level?

Are there other projects I should be supporting instead? I also have the old opteron 48 core beast in the basement. Is there anything worth doing with that these days?
 
Welcome back, 600k is a little low depending on the projects - a more usual number is around 750-800k ppd.

As for the 48 core opteron, there is a newer a7 core that can give reasonable PPD, i'm seeing around 270k PPD from a pair of ivy bridge xeons but it takes advantage of avx so i'm not sure what you would get. It might be worth a try and if not i'm sure the boinc team could use a hand.
 
jebo_4jc : put the 1080 on FAH and the cores on WCG would help both worlds. If life is easier your can also use the 1080 for GPUgrid ; an equal/similar project to FAH (one GPUgrid scientist now works for PG; world is small)

Whatever you do and as Gilthanis mentioned please put your cores on WCG for a while. Makes it nice warm in your room !
 
In that case, if you do decide to help out with WCG, I would recommend changing your preferences on the website to exclude running MCM work units as they can go up to 400MB and also exclude MIP which can take up to 650MB RAM each. https://www.worldcommunitygrid.org/help/viewTopic.do?shortName=minimumreq

If you ran a mix, I'm sure you would probably be ok, but since MCM has a higher priority setting server side, you may end up loading up on them. With only 16GB Ram, you would still need something for OS overhead too. 16GB was pretty common for BigAdv rigs.
 
Welcome back jebo. I agree with CV: 1080 on FAH and cores on WCG. I'm going to dedicate all my CPU's to WCG for the challenge myself. Crunch on!
 
Back
Top