Smp v6.34 was released about 3 weeks ago it uses the new A5 core, which is the only core that does bigadvs now, they stopped bigadvs on the A3 core about 2 weeks ago, upgrade, it's almost certainly the problem.
General Confusion, make sure you haven't accidentally turned the adv methods flag on in the client config.
I switched back to Linux this morning. Recieved project 2684 for my troubles but at least it's a step in the right direction.
I was just about to give you a shout to jump back in as I've had 3 linux bigadvs on the bounce, and my 3rd is a 2684
I've been picking up plenty of 2684s on the SR-2s. From the glass half full perspective, at least they're all bigadv now. This has been quite hit and miss. Some 6901s fold at a tpf of approx 19m16s whereas others chug along at about 19m42s. It's either one or the other with no middle ground seen so far. I'm not sure what's going on but the lower number kicked in half way through a work unit after switching to command line.
this is probably a bit of a noob question, but what the heck! is it possible/worth folding on GPU's aswell? as i currently fold using -smp 7 and 2 560's (@1GHz ) producing 56k-60k depending on WU's the 560's have. was therefore thinking i could use same setup in linux and get a slight increase in ppd due more efficient folding? Edit: by the way i am a complete linux virgin so any help would be grateful
Folding on GPUs under Linux can be done, but uses WINE to create a Windows environment for the GPU client. This is not officially supported by Stanford, but you can find out more by googling. So I think you are asking if Windows OS + SMP + 2x 560 scores more than Linux OS + SMP + WINE with (2x 560). I suspect the answer is 'yes' I've looked at the WINE stuff and decided it was not for me, another Linux newbie.
Folding on GPUs under Linux can be done, but uses WINE to create a Windows environment for the GPU client. This is not officially supported by Stanford, but you can find out more by googling. So I think you are asking if Windows OS + SMP + 2x 560 scores more than Linux OS + SMP + WINE with (2x 560). I suspect the answer is 'yes' I've looked at the WINE stuff and decided it was not for me, another Linux newbie.
Have you tried this: http://www.amdzone.com/phpbb3/viewtopic.php?f=521&t=138463 ? I installed it on my 970@4ghz Ubuntu, with the ck kernel. It was 17% through a p2686, each tpf was 20.07/8, after installed they were all 19.55/6, 12 secs improvement, not earth shattering but well worth doing, quick and easy to install, dropped it to the shell but it made no difference.
I gave it a bash at the weekend but thanks for flagging it anyway. It did wonders dropping the fastest SR-2 down to a 6901 tpf of 9m24s for a ppd of over 210K The ensuing excitement ended when the next wu dropped to a tpf of about 11 mins. A client restart set it racing again only for same sequence repeating on the next wu. I'll try to figure out what I'm doing wrong next weekend.
