New NVIDIA GeForce GTX280 Three Times Faster than HD 3870 in Folding@Home

[zekrahminator]

Just a couple days ago, we informed you that NVIDIA had joined the Folding@Home team. However, at the time, benchmarks for this new client were unavailable. I am now happy to inform you that (internal) benchmarks are available for your viewing pleasure. The rather large green bar was achieved using the new NVIDIA GPU core, the GTX280. As far as exact numbers go, this sucker can fold at 500 mol/day, which is much higher than the Radeon HD 3870 numbers (170 mol/day), five times higher than PS3 numbers (100 mol/day), and astronomically higher than the average computer numbers (4 mol/day). Whether or not this translates into actual gaming performance is yet to be seen, however, it's pretty hard to imagine how something so powerful wouldn't bring back some respectable FPS in games like Crysis.






View at TechPowerUp Main Site

 

[ShadowFold]

Is it actually called GTX280 or is it 9900GTX?

 

[panchoman]

as i said in the other thread.. i'm calling BS on this graph

 

[Darkmind]

what's "mol" ?

 

[magibeg]

Well that would explain things a little bit better from that Folding@home slide before when it showed the nvidia gpu killing the ati one. Also makes me curious as to what a 4870x2 could do in terms of folding.

 

[jocksteeluk]

Great! how does it do on games?

 

[DrPepper]

Hurry up goddamit I want to fold my 8800GT. I hope this client allows other cards to fold and not just the GTX280

 

[Animalpak]

Sad years for ATI ( OWNED AGAIN ).

 

[Para_Franck]

Owned by speculations.... Com on, we don't even see 48xx performance....

As I always say: wait and see.

As for the 38xx series, they perform verry well for their price. They are beat by Nvidia, but not owned.

 

[panchoman]

benchmarks show that a single 4870 pwns a 9800x2 in crysis and 3dmark

 

[Animalpak]

Will see this summer.

 

[philbrown23]

now lets compare it to the 4870 then we will see whos on top

 

[DrPepper]

Sad years for ATI ( OWNED AGAIN ).

Its far too early to make assumptions because neither cards have been released or even benchmarks released.

 

[Azn Tr14dZ]

what's "mol" ?

It's a counting unit. One mole (or mol for short) is 6.02214x10^23...which is Avogadro's Number. I just learned this in Chemistry.

 

[Para_Franck]

What strikes me most is how the PS3's cell processor is easily beat.

 

[WarEagleAU]

Considering all the time Nvidia has had to work with the folding at home code and finally get it going, Im not much dismayed by this. Im glad they can fold that many molecules a day myself, this will help us in the future no doubt. also, considering what their CUDA platform can do, this isnt a huge shocker; even though ATIs GPGPU is pretty awesome in its own right.

PS3s cell processor assists in the folding if Im not mistaken, but its actually the nvidia part that does most of the work. I of course, could have that backwards.

 

[acperience7]

Great! how does it do on games?

My thoughts as well. For something that is basically a side note feature Nvidia is making too big of a deal out of this.

 

[farlex85]

I wouldn't doubt it, this card will likely be a beast. Probably the best money can buy for a while.

 

[Azn Tr14dZ]

Considering all the time Nvidia has had to work with the folding at home code and finally get it going, Im not much dismayed by this. Im glad they can fold that many molecules a day myself, this will help us in the future no doubt. also, considering what their CUDA platform can do, this isnt a huge shocker; even though ATIs GPGPU is pretty awesome in its own right.

PS3s cell processor assists in the folding if Im not mistaken, but its actually the nvidia part that does most of the work. I of course, could have that backwards.

Wait, so does F@H fold moles or molecules?

 

[DrPepper]

Wait, so does F@H fold moles or molecules?

Maybe the moles of the molecule

 

[panchoman]

well theres avogadro's (6.02x10^23) number of atoms/moles/molecules/particles(the most generic term) in a mole (also 22.4 L at STP)

so F@h analyzes moles of molecules, making dr. pepper's joke true lol

 

[imperialreign]

just thought of something . . .


well, any of us who've taken part in the F@H project know that each console is otpimized for what it's supposed to be working with - they literally pick the work load for each specific console which is optimized for specific hardware to run it best


so, seeing as how SMP work units are optimized for multi-core processors, and CPU clients are optimized for single-core processors . . . the GPU client we've had for our ATI cards runs WUs that are optimized for a GPU, and the PS3 client runs optimized WUs for it's processing architecture . . .


my thought, then - what's to say that nVidia's graph isn't just comparing what their unit IS capable of, but incorrectly compairing different types of processing - sure, their new hardware might be able to process 500 molecules a day, but that doesn't really mean much in the overall scheme of things if it's processing simple molecule WUs instead of more complex units like the SMP client can handle . . .


. . . not sure if I'm making any sense with this, I feel like I'm rambling

 

[DrPepper]

well theres avogadro's (6.02x10^23) number of atoms/moles/molecules/particles(the most generic term) in a mole (also 22.4 L at STP)

so F@h analyzes moles of molecules, making dr. pepper's joke true lol

It was true I learned about moles in chemistry too and my exam is on Friday

 

[panchoman]

just thought of something . . .


well, any of us who've taken part in the F@H project know that each console is otpimized for what it's supposed to be working with - they literally pick the work load for each specific console which is optimized for specific hardware to run it best


so, seeing as how SMP work units are optimized for multi-core processors, and CPU clients are optimized for single-core processors . . . the GPU client we've had for our ATI cards runs WUs that are optimized for a GPU, and the PS3 client runs optimized WUs for it's processing architecture . . .


my thought, then - what's to say that nVidia's graph isn't just comparing what their unit IS capable of, but incorrectly compairing different types of processing - sure, their new hardware might be able to process 500 molecules a day, but that doesn't really mean much in the overall scheme of things if it's processing simple molecule WUs instead of more complex units like the SMP client can handle . . .


. . . not sure if I'm making any sense with this, I feel like I'm rambling

i see what you're saying there imperial, and it think you've struck something here

i'm wondering what kind of core the wu thats used in the graph based on? is it even based on a wu? is this just processing power? or is this theoretical? what is the basis for the data for other things? what wu's are the other things in the graph running?

makes me think this is some theoretical processing power BS

 

[DrPepper]

It looks too exact (500) to not be theoretical.