PPD per graphics card/Processing core

[Englander]

The GPU client has a pause work option on right clicking, and the smp console window can just be closed when you want to pause it, it will save and restart where it left off on next running.

Indeed, that's the system I have now adopted

[Tattysnuc]

A nice little acticle here on Bit-tech regarding folding performance.

http://www.bit-tech.net/hardware/gra...-for-folding/1

Key thing that I've noted: If you've got a 200 series card stick with GPU2 as GPU3 is 44% slower

Wow! I'm gonna check what my GTX 295 is folding on when I get home tonight. That's only pulling in 11k a day vs their projected 15k

[Queelis]

Yeah, just the thing I've read just now. It's frustrating that ATI is so hopeless at folding, it's hard to believe it

[Tattysnuc]

My ppd has dropped somewhat on my second rig (ITX board with a Q6600 @stock, with a GTX 285 @ 680/1760/1252) recently from a toasty 9900 ppd to a meagre 3800 ppd according to fahmon. It's definitely running slower, or rather taking longer to complete work units.

Does anyone know if this is right or if this is due to the kind of work units out there or soemthing else? I've rebooted the machine which has fixed this in the past, but recently it's dropped right off.

I'm still on the v2 of the GPU client, so it's not because of running the newer slower client.

Ideas anyone? The machine is running fine otherwise....

[Tattysnuc]

...Odd. For the best part of a week the card folds @ 1-3k, then the moment I post on hexus, ppd climbs to 8k again.

Now, what can I write to make me win the lottery....

[Queelis]

Are you sure you haven't been doing any gaming on those? Otherwise, maybe it's down to the different projects having different peculiarities.