Folding Chit-Chat

[Golden Dragoon]

I have always pronounced it Hex Us, something about that pronunciation hurts my ears lol

[Tattysnuc]

Fold for Hexus!
A possible recruiting tool, no?

That's going straight into my signature! Genius!

[Englander]

Just tried installing the SMP client on my laptop (don't expect any -bigadv units!) and it's only using an average of 25% of the CPU/ Set the -smp flag both in the config and in the shortcut... any suggestions?

EDIT: deleted the work folder/queue.dat and re-ran it, it's up to 100% now

[Tattysnuc]

I've had to stop my GTX 480's folding on my main rig. Because I had a couple of BSOD crashes I was going over the deadline for the first BIGADV submission.

I've now overclocked the 970 up to 4.4Ghz on all 12 "cores" using a meagre 1.35v and it's going great guns. 22-25mins per fold.

Unfortunately it looks like I'm going to be 30 minutes over the deadline! If only I'd spotted that this morning. There was definitely something not happy with it at 4.25Ghz. Must have been the memory strap or something, but it definitely ran slower. As Fermis are known to gobble "arfa" core per client, maybe I do need to sit down and set up the clients more efficiently (Cue Queelis!)

But for now.....DOH!!!!!

[Englander]

Actually running -smp on my laptop isn't good for the ol' sperm count :/

[1stRaven]

I've had to stop my GTX 480's folding on my main rig. Because I had a couple of BSOD crashes I was going over the deadline for the first BIGADV submission.

I've now overclocked the 970 up to 4.4Ghz on all 12 "cores" using a meagre 1.35v and it's going great guns. 22-25mins per fold.

Unfortunately it looks like I'm going to be 30 minutes over the deadline! If only I'd spotted that this morning. There was definitely something not happy with it at 4.25Ghz. Must have been the memory strap or something, but it definitely ran slower. As Fermis are known to gobble "arfa" core per client, maybe I do need to sit down and set up the clients more efficiently (Cue Queelis!)

But for now.....DOH!!!!!

Thats going well for the cores. My 24 core servers are running around the 20 minutes per fold on the bigadv units - granted they are running at only 2.667 and 2.550Ghz.

[Tattysnuc]

I've killed the GTX 480's folding. Looks like they are interfering with the production rate of the BIGADV. If that is so I'll just leave the CPU folding.

Had to back it down from 4.4Ghz as it bluescreened after an hour. Looks like I should be able to get it to fold at 4.4-4.6 Ghz with enough juice, but TBH I'm not sure that I want that much stress on the board as it is my main gaming/work/stuff rig.

I've popped it down to 4.2 and looking at the overnight stats with 2xGTX480's running flat out @ 785/1570/2000 on each it looks like they were struggling with those overclocks too. I backed down the overclock to 750/1500 and left the memory @ 2000 and the GPU usage went back up to 100%, whereas before it was fluctuating about the 80-90% mark.

The CPU seems to be plodding along now, between 65C and 70C (1.34v Vcore) which I know I can get down if I upgrade the waterblock to my XT, but that's currently in bits missing a couple of O-rings.

With everything running, the BIGADV was projecting to finish in 5 days, which is taking far too long per fold (1.5 hours!). I've knocked off the GPU's and set the priority to HIGH on the A3.exe and the main client to see if it's them affecting it and I'll have another look tonight when I get home from work.

Fingers crossed that's what it is. I'd read somewhere on Bit-tech that Fermi's require more setup with the CPU's as they require half a core per GPU. If that's the case I'll have a go at setting the affinities over the weekend, and only folding on 11 threads instead of 12 for the BIGADV.

[watercooled]

Slowly climbing my way to the 10th WU! I keep looking for more info on how to monitor Thuban temps, using programs like CoreTemp with stock settings report the CPU at about 8-10C idle and 25-27 running f@h. The best advice I've found is to add 10-15C to the reported temps to get the real values, which would give C25C idle and about 40C which sounds about right. Does anyone have any better advice?

[scaryjim]

Woo! 250k

I'll think about firing up the 8800GTX again towards the end of the month to push towards the 100M - I know it doesn't up the team by much, but every little helps right?

[1stRaven]

Woo! 250k

I'll think about firing up the 8800GTX again towards the end of the month to push towards the 100M - I know it doesn't up the team by much, but every little helps right?

Certainly does. I fired up another Bigadv server today to run over the weekend, mainly to try and keep ahead of Tatty....

[Englander]

2 days away from the 2 million mark Plodding along at about 4 GPU WUs a day so ~5k PPD.

[shbris]

what gpu is that? ur sig says 260 but i would have thought that would do more PPD than 5k

[1stRaven]

Keep it up. Only 2 weeks off the number 1 slot here but Tatty's new rig is going pretty well. He should be over the 100kppd mark soon.

[Englander]

what gpu is that? ur sig says 260 but i would have thought that would do more PPD than 5k

It's a 460 so in theory capable of 1.5-11k PPD depending on the overclock, but I no longer have free electricity so I've downgraded to part time :/

[shbris]

that's fair enough, folding is a leccy bitch but i run it 24/7 in my room instead of radiators being on so i reckon it all works out ...

[watercooled]

I can't seem to find a simple answer to this with Google - do you get bonus points for the normal and GPU clients or is it SMP only?