Whoopass !!!!

[Marvin_The_Martian]

No serieusly I mean it... WTF

Not to mention OMFG

Anyone got a stretcher? Those be a big nvidia farm or what lol... would be nice manufacturer surprise, we had such a large stock off g92 cards we had the idea to set up a small folding farm in our headquarter as promotional material. They will run indefinetly from now on...

Keeps dreaming...

Edit:


http://www.tgdaily.com/content/view/37931/113/

And no one tells me lol, 8 people look and wonder where the hell I been the past few weeks ( and surprisingly no one says I should search more? ) I mean this has been posted here right?

[Sparky]

well nvidia did say they were gonna open a big can a little while ago....

[Marvin_The_Martian]

Yeah but why is everyone so about this.. I posted in the news section just now after looking through Dinos his thread and not really finding any references to this teams output. Thought it deserved some more attention don't you think? And damn I wonder how many cards are running under MacFolden

[MikeB12]

I discussed them here a little a couple weeks ago... http://www.xtremesystems.org/forums/...light=whoopass

but they were new, only folding on gpu2 for a day and a half at the time...
they've cracked up the ppd since then...

here's the team mentioned in the article... nvidia team whooopass... they have been folding for 2 days and have 400K...
so they are getting 200,000ppd on 4-5 test machines if the OP article info is correct.
http://folding.extremeoverclocking.c...hp?s=&t=131015

[Marvin_The_Martian]

I discussed them here a little a couple weeks ago... http://www.xtremesystems.org/forums/...light=whoopass

but they were new, only folding on gpu2 for a day and a half at the time...
they've cracked up the ppd since then...

1% off all CUDA enabeld cards people, just one percent and Vijah will have the fastest supercomputing cluster ever. Now, I'm an optimist and hope we can get everyone, and I mean everyone, on this and wth force nvidia to incorperate folding in their drivers

Well ok not force upon, but everyone should know in their pc might lay the power to cure cancer for everyone including their own loved one's. Fah was already big, I think f@h now can be enourmous. The best example of what DC and a united community can do

[ThunderDuck]

Doing my part.
I'll be adding an Ultra KO and GTS (G90) this weekend.

[Marvin_The_Martian]

What cpu's you're pairing them with? I'm finding that under vista x64 the gpu client shows only 3% usage will still doing 4k which is normal I guess for my card and clocks, so my smp hasn't gone down much if at all. Which is strange when you read it needs an entire core, but it's folding I think 100% with just 3% cpu usage.

[ThunderDuck]

The Ultra is going to be on a E8400 @ 3.4 and the GTS is going with a X3350 @ 3.2
I'm going to kill one of the smp's on the 3350 and give the GTS one core and the other smp will get three. Both are running XP.

[Marvin_The_Martian]

Well like I said try vista x64 I have a ss in a thread here somewhere showing fahmon taskman and rt all running and I tell you it seems to take just 3% from one core on this os for some strange reason :o Both might still run smp and gpu, though I would see what the clients takes when the machine is iddle anyway. Mine never goes above 3% ever for some reason and 4K from a 9600gt seems to be pretty good.

[coo-coo-clocker]

I'm running:
8800 GTX stock speeds
F@H GPU2 6.12b6
Nvidia forceware 175.16

For project 5004, I'm getting:
Avg. Time / Frame : 1mn 39s - 4180.36 ppd

However, it's eating about 50% of my CPU (split across both cores)!

something about this doesn't seem right. I am using the "slightly higher" core priority setting in F@H...

[Marvin_The_Martian]

That's not good for sure, I get 39xxppd on that wu on a 9600gt. Show us a taskman shot please, want to know the processes which are running where

[angra]

coo-coo: can you verify that you are running the most up to date client?

[Marvin_The_Martian]

Also now that I think about it, are those drivers officially supported ( I think not )?

[coo-coo-clocker]

I upgraded to the beta 8 version of the GPU2 client.
No joy - still gobbling half my cpu.
The taskmanager shows the fah core exe running at 50% utilization.
Don't know about the highest supported forceware driver version though. I thought that I was good on that...

[SocketMan]

Couiple of days ago somebody said something about :the client is beeing fast,
I replied by saying :"It's a little too fast, we're about to get our ass whooped
check out the team stat page.
Maybe if we buy all the gtx280 in the world that would slow the whoopass
down?

[bowman]

A few Tesla S1070+quad socket racks. And a micro-fusion reactor along with them.

[dinos22]

Yeah but why is everyone so about this.. I posted in the news section just now after looking through Dinos his thread and not really finding any references to this teams output. Thought it deserved some more attention don't you think? And damn I wonder how many cards are running under MacFolden

amazing stuff isn't it

[Movieman]

I read that comment of "163 times the processing power of a quad core cpu"
WOW!

[RottenMutt]

i need to go out and get one, within a week i could pass you up mm

[Movieman]

i need to go out and get one, within a week i could pass you up mm

If you meant me, that would be easy. I'm not running F@H but I support you guys totally. I was just amazed when I saw that 163 times figure.
dear Lord thats some spicey meatballs..

[MikeB12]

Pande came back and said those numbers were exaggerated in Tom's article... so I don't know whether I'd trust the 163 times comment... but the new gpu2 definitely kicks the quads butt...

remember also the cpu wu's work with a lot more atoms.
for example:
a smp wu 2665 works with 14170 atoms...
a gpu wu 5001 works with 576 atoms...

so it's hard to compare them equally..

[Sir_Loin]

Pande came back and said those numbers were exaggerated in Tom's article... so I don't know whether I'd trust the 163 times comment... but the new gpu2 definitely kicks the quads butt...

remember also the cpu wu's work with a lot more atoms.
for example:
a smp wu 2665 works with 14170 atoms...
a gpu wu 5001 works with 576 atoms...

so it's hard to compare them equally..

Sooo true. I'm waiting to see if the pointage holds this same line in the weeks to come, or if we'll see some adjustments by PG.

Regardless, team whoopass is doing just that, and even though our teams will probably be overtaken, the project is benefiting as a result.

[[XC] Synthetickiller]

Wow, I wish I could let my computer run 24/7 considering i'm supposed to be getting 5600 to 6000 ppd (the figure jumps), but sicne I leave my computer on 12 hours usually, i submit about half that.

Btw, I ham running the beta 8 GPUv2 client and I take NO cpu hit, very little if anything. I was able to fold while running prime 95. My iterations did not drop by much if anything. It doesn't hurt that i'm running a quad, but still. I'd love to build some e7200 machines and run a few gtx260s or 280s on there, but i have 0 cash flow and my next build will probably be a 6 core nehalem (and that's in 2011), so for now, I'm SOL on building more systems for a while.

Is anyone running the GPU & CPU client at the same time? Also, Has anyone tried running the nVidia and ATI client at once? Is that possible? That might allow people to do multi GPU folding w/ only one computer.

[bowman]

Btw, I ham running the beta 8 GPUv2 client and I take NO cpu hit, very little if anything.

Are you on Vista? The CPU utilization problem seems to be affecting XP users mostly.

[[XC] Synthetickiller]

Are you on Vista? The CPU utilization problem seems to be affecting XP users mostly.

Perceptive young grasshopper! Thanks for the heads up, I didn't know that. I guess vista did one thing right!