Let's go Team !!

[LCB001]

NIce scores. For some reason, on that project, my GTX 295 is showing brutally low PPD.

I'm going to wait until the morning to see if it continues like this, but it's showing about 3k ppd per GPU at the moment whereas before it was showing more like 6k per.

I can't find it right now but I've read posts about some of the 200 series cards not liking those WU's. Each half of your 295 should at least equal a 260-216, my GTX+'s do ~6014 ppd on those WU's.

@ chrisk; Your GSO might be a little low, the GTX+ looks good. My GTX+ does ~6014ppd but my shaders are at 1998 MHz and core/mem came pre OC'd which might help a little...

 

[zlojack]

Woah thats pretty low. On the same project I am getting ~8800PPD on my 285s. You should be getting ~6000-~7000 on that per gpu no?

I figured it out. The VMs were running at "Normal" priority and taking up all the CPU cycles. I set the VM priority to "Low" and now the PPD on the GTX 295 is back up to around 12.5k so far :biggrin: and I'm sitting at about 24k total :clap:

I can't wait to get this whole rig overclocked!

 

[chrisk]

I believe my GTX+ is OC as well (770 core/1125 mem/1944 shader). Shaders won't budge one single bit, which may be just a card at its limit or some variation of the shader bug. I also think I can push the GSO a little bit, but I got a bunch of of 511s so I got scared.

Probably will be replacing that GSO in a week or two, and pop in a GTS 250 to pair up with the GTX+. I'll do a fresh OS install with new drivers, and see if anything happens with my GTX+ (better/worse PPD, better shader OC, etc)

 

[chrisk]

Oh dear...I only played with VM...am I missing out on some PPD? I am running the smp on the folding rig in my sig, and nothing at this moment on the Q6600...

 

[LCB001]

Oh dear...I only played with VM...am I missing out on some PPD? I am running the smp on the folding rig in my sig, and nothing at this moment on the Q6600...

Using a VM and the linux SMP client would boost the PPD by half at least since the linux client can run the A2 SMP core. netforce linked a nice guide to getting it set up here; Core i7 Folding - how to increase your PPD - Xtreme CPU. It will work for any set-up not just the i7. It sounds simple enough I might even try it on my Q9550, just make sure to read the whole thread.

You should definately get your quad going, running two VM's it should be good for ~4500+ ppd at least...:thumb:

 

[LCB001]

Breaking the 100 mark is absolutely amazing considering the size of the team, and the fact that we're up against teams that have been folding for 5+ years.

It says a lot about the dedication of our members that we manage to put out the points that we do. Good Show to all!!!!!!!

Almost missed this. If it wasn't for the dedication of long time members like yourself and 3.0charlie and some of the others that kept it going we would't have a Team, so Thank You...

 

[zlojack]

Using a VM and the linux SMP client would boost the PPD by half at least since the linux client can run the A2 SMP core. netforce linked a nice guide to getting it set up here; Core i7 Folding - how to increase your PPD - Xtreme CPU. It will work for any set-up not just the i7. It sounds simple enough I might even try it on my Q9550, just make sure to read the whole thread.

You should definately get your quad going, running two VM's it should be good for ~4500+ ppd at least...:thumb:

I used this guide from the EVGA forums as well as the one you linked and it was really easy!

I wish I'd done it sooner.

 

[LCB001]

I used this guide from the EVGA forums as well as the one you linked and it was really easy!

I wish I'd done it sooner.

Now that some of these guides are starting to be written in actual English it's harder to find reasons to put off doing it, as long as it's as stable as my windows SMP's the extra PPD is just too tempting.

I just wish Stanford would hurry up and get the new Windows SMP client going, yes I know, good luck with that...:blarg:

 

[zlojack]

Now that some of these guides are starting to be written in actual English it's harder to find reasons to put off doing it, as long as it's as stable as my windows SMP's the extra PPD is just too tempting.

I just wish Stanford would hurry up and get the new Windows SMP client going, yes I know, good luck with that...:blarg:

Yeah...seriously. Though I understand why they focus on the GPU client since it does crunch the numbers very quickly.

Still, the A2 for Windows would be very nice :biggrin:

But now that we have a guide that has everything including PICTURES, there really is not very much of an excuse not to go the VMWare route (yes, I'm looking at you, 3.0charlie :haha

 

[chrisk]

Getting about ~800PPD on the 6400+...I'll let ya know how she does with the virtualized version.

Great guides BTW and much appreciated. I noticed something in the evga guide though...in step 2, they include the client no? Or did I miss something? Is that not against Stanford's TOS? I won't complain, as it makes it easy, but I thought that F@H had to be downloaded from Stanford directly?