Quote:
Originally Posted by geokilla Did you post over at the Folding Forums about the problems you're getting with Project 2684?
Is your RAM stable? Temperatures in check? |
On my most reliable two i7 920s I've had them fail at startup, and on my one i7 860 it failed at 63%.
All three boxes have been memtested and Prime95/Linx stable tested for 24 hrs and this was while the GPU clients were running to make load as realistic as possible. Ie. Prime95 or LinX and GPU folding at the same time.
All three at their current settings have never failed an A2 BigAdv unit unless a power outage happened.
Also none of them exceed 70 celsius on any core while SMP folding or BigAdv A2/A3 folding. One, an i7 920 C0 that can do 4.0 Ghz at its current voltages, is at 3.8Ghz and is super stable, and yet has also failed two or three A3 BigAdv units at startup.
Quote:
Originally Posted by chriskwarren Other folks over there at FF complaining about A3 units too. They are the bigadv equivalent of the  511s. |
I happen to be one of them along with some guys from [H]ardOCP. I posted my logs to show them the issues. We'll see what happens. This better improve or my last two BigAdv folders will be WinSMP converts.
Quote:
Originally Posted by Rison I know what you're talking about 10e - the other reason I asked, is because I had so many units fail on my 860 box.. about three in a row when they started getting the a3 units.. That, and it takes basically 4 days to complete a work unit now. Longer time = more chance something will go wrong, especially with these new units.
I just had both my i7 bigadv boxes fail their workunits last night, for reasons unknown - after about a day in of folding each. |
Yeah, that's horrible. We can't cherry-pick the BigAdv WUs. WinSMP is looking pretty sweet now even with the PPD hit.