6th Annual Folding@Home Holiday Season Race

[theAnimal]

I have an important issue:

on win7 latest drivers and client 7 two of my nvidia gpu computers freeze randomly in the middle of gpu WUs...

anyone else having this problem? what could be wrong, it even happens without OC (the OC was gpugrid stable for month anyway)

Try older drivers, or even underclocking (especially RAM since it makes virtually no difference for F@H).

 

[petrusbroder]

@theAnimal --- hooray!

 

[GoStumpy]

I'll bring my lowly i3-2100 on for the team, wondering, it only shows 25% CPU usage... am I doing something wrong?

 

[geecee]

I'll bring my lowly i3-2100 on for the team, wondering, it only shows 25% CPU usage... am I doing something wrong?

Which version of the client software are you using and are you running the SMP client? It seems like you might be running the uniprocessor client, as 25% usage is consistent with using just one out of the 4 threads that you would have available with the i3.

 

[GLeeM]

I'll bring my lowly i3-2100 on for the team, wondering, it only shows 25% CPU usage... am I doing something wrong?

If you are using 25% I am going to guess that i3-2100 is dual core with HT and that you are running the Uniprocessor client.

The SMP client will use all four cores to finish work faster for a bigger bonus.

 

[theAnimal]

@theAnimal --- hooray!

I've got another one finishing in less than 28 hours.

I should exceed my last year's race output on Tuesday or Wednesday at the latest.

 

[petrusbroder]

Here are the current stats:

Another big dump by theAnimal ... what is he folding with? A super computer farm? The borgh-mothership?
Most folders have increased their production. That is very good. Today the recers produced som 75%+ of the TeAm's total.
However: there are som absentees? Anybody know something? Has anybody PM'ed Mark (Assimilator1) and Sid (Insidious) ... don't we want to increase the teams? Get on recruiting, asslaminating, borghing ... resistandce against the 6th Annual FoldingHome Holiday Season Race is futile!

 

[somethingsketchy]

I should probably throw out there, that I've been folding under the tag *Ryan_B* for the last couple of days.

 

[petrusbroder]

Oh, BTW, here are the stats from the 5th Annual Folding@Home Holiday Season Race:

We should beat that, shouldn't we?
More folders to the race, more cores, more GPUs, more TeAm-mates, more, More, MOre, MORe, MORE!

 

[petrusbroder]

I should probably throw out there, that I've been folding under the tag *Ryan_B* for the last couple of days.

Good to know ... I'll include you in the stats from tomorrow. :whiste:

[BTW: different names are very common in F@H. Just look at the table ... ] :biggrin:

 

[Rattledagger]

Another big dump by theAnimal ... what is he folding with? A super computer farm? The borgh-mothership?

He's using the Animal-farm of course. :whiste:

Great production again today.

 

[petrusbroder]

LoL!

 

[Pokey]

With the new V7 software, is it necessary to set the "advanced" & "big packet" flags as we do with the older software. (I'm talking about a quad core Q6600 here.)

When I tried to set those flags for the GPU it gave me an "invalid argument" message.

Trying to learn my way around the V7 Client.

 

[somethingsketchy]

Personally I've been using this as my F@H software and so far it's been working like a charm.

http://hardforum.com/showthread.php?t=1550668

Basically you fill out a few options (team #, your username, passkey, etc) and the software will download a new WU. Right now the software will handle CPU (single core), SMP (multi core) and eight GPUs. You can also enable -bigadv, -advmethods, etc. So far I haven't had any issues with my main rig on four threads, and it's completely six WUs in the last couple of days.

 

[Markfw]

So, I have moved teams ?

 

[Rudy Toody]

So, I have moved teams ?

Or maybe you stayed put and they moved.:whiste:

 

[GoStumpy]

Question...

I have 3 or 4 spare computers (P4, Athlon 2800+, 3000+, etc) that I could just leave running in my spare room.. however they won't have access to the internet... I have a single wireless USB stick... is there a way (without checking on them constantly) to just have them store packets, work on them for a week, then plug the wireless into them and upload?

I used to do RC5 crunching many years ago, this sounds like fun

 

[Pokey]

So, I have moved teams ?

Peter also posted last years stats chart. You are probably looking at that.

 

[Markfw]

Peter also posted last years stats chart. You are probably looking at that.

Yup, that was it....

 

[theAnimal]

Another big dump by theAnimal ... what is he folding with? A super computer farm? The borgh-mothership?

Dual hexcore Xeon @ 4GHz. And my daily driver hexcore i7 @ 4GHz.

 

[petrusbroder]

@GoStumpy: What is your name while Folding? I have not been able to find you in the stats!

 

[GoStumpy]

Might as well use this double post... lol

http://fah-web.stanford.edu/cgi-bin/main.py?qtype=userpage&username=GoStumpy

Hurray I'm on the board!

 

[GoStumpy]

Just started today, perhaps I don't show up until I've submitted a WU or two... I used to fold for vwvortex.com under DasCoupe.

http://fah-web.stanford.edu/cgi-bin/main.py?qtype=userpage&teamnum=116349&username=DasCoupe

 

[ZipSpeed]

Showed up to work and saw one rig off from a power failure. Three other rigs were functioning but couldn't get any work and had to power cycle the switch which I can thank again to the power failure. The last three rigs manage to survive the power failure somehow and continue to crunch happily.

Anyway, I'm back to full capacity but looks like my PPD will take a hit in the last 24 hours.

 

[blckgrffn]

Sooo.....

After spending too much tweaking and installing and crap to get onto an OpenCL version that works (I think I could have actually stayed with Cat 11.11...) my TPF is now ~14 minutes versus 5 and it looks like the current WU is going to take nearly 24 hours versus the ~7 hours it took before. OpenCL can't be slower, can it? How can verify that I am using OpenCL?

The log is showing AMD/OpenCL and the GPU type and in the visualization window it lists openmm_opencl as what is running.



Although...

Previously I was working on a 550 point project, 5726 or something and now I am on a gigantic 1835 point monster (in my limited experience) - so maybe it was using OpenCL before and I just spent this time for no reason? lol. I guess I should have looked into it more before changing things - although I could have sworn the visualization thing said something about a stream before - what is the "old" AMD/ATI message?

The GTS 250 is out for now - "unstable machine" responses every couple of WU's. Everything is stock clocks and it happens when it is trying to load a new WU, it always finishes and uploads WU's. Must be invalid work?