ready for some Fermi numbers? GTX 480

[Otis11]

crap. only 30% faster than gtx285, but price difference is more than 30%

But should soon be 100% faster.

[lkiller123]

I only have one basis for equating the 480 for CUDA ... on Collatz sftp (the major dude who contributed a 470 and a 480 to GPUGrid) said his 295 would crunch a WU in 18minutes. My 480 is cranking them out in 7 minutes. That is not to say GPUGrid will do better or worse in optimizations but as the only piece of evidence I can point to it looks pretty good to me. Lastest post from GPUGrid regarding fermi is they expect to have a linux version ready tomorrowthey are now focused on optimizations and they may have a Windows version ready also. Hold on to your hats, I'm gonna put up some serious numbers (well for me anyway )

The guys at GPUGrid really do move along quite quickly. If you follow all the different tech they try out and the papers they publish, top it off with the public project ... they are Xtreme

That seems like good news But are you sure that his 295 crunches a WU in 18 minutes in one of the two GPUs or both of the GPUs? I don't really know how Collatz works, though

[trn]

crap. only 30% faster than gtx285, but price difference is more than 30%

And I woulda have bought a few of the output was more of what we were expecting.

[=[PULSAR]=]

And I woulda have bought a few of the output was more of what we were expecting.

They are still tweaking the issues. They discovered they have to rewrite a lot of things because everything was based on the older architecture. The 30% increase was with using Fermi on the older GTX2xx architecture.

From GDF

Current performance on the GTX480 is 35% faster than the fastest version of ACEMD on a GTX285.

There is little more we can do I think. The cuda3 compiler seems much worse at optimizing the code than the 2.2 version. If this is the case, then 3.1 could bring some improvement.

Tomorrow, we clean up and submit the new application in beta.

gdf
PS: no doubts this is much less than expected

[stoneageman]

Very disappointed in the performance figures quoted by GDF Looks like the 295 will still be the king of cards for a while. Glad I waited before ordering now

[WhiteFireDragon]

yes, having blown over $1k on two of these, i sort of regret spending on so much. i was just so excited when i saw it in stock. well to minimize buyer's remorse, i'm going to be optimistic and know that they'll be optimizing the WU's for these cards within the next few weeks

[=[PULSAR]=]

Update

we have now a good performance model of the application. The maximum we should have expected from fermi is a reduction in time of 60% compared to a gtx275. Actually, we now think that this will be achievable in the future. The problem is that cuda 3 is slower even on gtx200 by almost 15%. Indeed, we are 60% faster compared to both applications compiled for cuda3 but compared to cuda2.2 only 35%.

On our machine the running temperature is 91 degrees. Equivalent to a gtx275.
The only real problem is the price. Later i will report the time of the gtx470 and running temperature.

Gdf

[INFRNL]

Update

if this is true, then fermi is not worth the cost, right? just trying to figure out if my head is working correctly.
if the 480 is max 60% faster, 60% more cost would mean the 480's should only cost $400

[YukonTrooper]

crap. only 30% faster than gtx285, but price difference is more than 30%

Well, that's hardly ever not the case with new hardware.

Although, to be fair, Fermi was looking amazing for math. I still think the potential is there, though. How can it not be with those @Home benches?

[123bob]

guess this gives me a little time before i bite. no sense it getting one if we can't fold yet. although i need to sell off another gtx275 or 2 before those lose value an i can use that money to help fund the 480. i hope a 480 can produce well more than 2 275's can otherwise would not be worth the investment

PM me. I didn't think you had any left.

bob

[=[PULSAR]=]

Okay well fermi is up and running now just waiting on the gpugrid to release beta they said sometime tomorrow.

[Snow Crash]

@Pulsar ... have you run any Collatz to get a feel for your OC? It runs a higher on the GPU utilization than GPUGrid so I think we will be stable if we max Collatz. Nice way to get a feel for the card in general, it's OC capabilities and the temps. Currently I am rock solid at 1672 but can push it to 1732 at stock v. I have not had much sucess OCing higher with more volts but I am sure I will be working on that again once GPUGrid is ready for us.

[=[PULSAR]=]

I'll give it a shot later on this evening trying to hammer out the last kinks in my WCG farm, almost had it finish and then HDD died. Back to using a prehistoric IDE

[lkiller123]

Isn't it a better idea to buy used GTX295s off eBay? Some of them are going under 300 dollars already.

[=[PULSAR]=]

Isn't it a better idea to buy used GTX295s off eBay? Some of them are going under 300 dollars already.

I have 2 gtx 295's and you can't run more than 1 per machine on evga MB's for gpugrid so byebye 295's

[lkiller123]

I have 2 gtx 295's and you can't run more than 1 per machine on evga MB's for gpugrid so byebye 295's

Is it only for EVGA MBs? I am thinking of getting some used 295s in the summer. I bet their price will fall like crazy by summer

[=[PULSAR]=]

Is it only for EVGA MBs? I am thinking of getting some used 295s in the summer. I bet their price will fall like crazy by summer

Yeah tried it on both E760 and E758 tried it on an older s775 and AM2 with no issues. I'm looking at getting rid of mine shortly probably within the next month or two. Want me to contact you when I do?

[lkiller123]

Yeah tried it on both E760 and E758 tried it on an older s775 and AM2 with no issues. I'm looking at getting rid of mine shortly probably within the next month or two. Want me to contact you when I do?

It will be no earlier than Mid June/Early July before I have time to do builds. You know how fierce high school is I will be happy if you do contact me, though

[WhiteFireDragon]

just got mine today and been really frustrated with it . look at all these errors it's giving. i've been trying to fix it for over an hour already and still no luck.

it seems that nvidia only has one driver to cover all cards except for the new gtx400 series, which has it's own drivers. i updated it, and still doesn't work. it keeps going through a loop of downloading the WU, and instantly report it back as error, and DL's a new one, and same thing happens. any idea how i can just get it running or is fermi still not yet supported?

[=[PULSAR]=]

GDF says it is beta is supported for linux. He says it may or may not work on windows yet, some users are getting it to work on XP. I tried running it today on win7 and no luck.

[WhiteFireDragon]

are you getting an infinite loop of error as well? i'm not sure if BOINC even detects mine. this is on win7

[=[PULSAR]=]

yeah running win7 64 here and as soon as a unit start it errors and loads another.

[Chumbucket843]

maybe you could for the time being run gpugrid on linux using vmware?

[=[PULSAR]=]

I'm currently to addicted to the supersonic sled demo haha

[Snow Crash]

yeah running win7 64 here and as soon as a unit start it errors and loads another.

That was happening to me also on Win7 64 ... then I turned my OC down until I started processing properly. I think I'm at high 15XX and stable.
I also set the environmental variable they posted (SWAN_SYNC=0) which forces the GPUGrid app to use a full core per WU, they say you get a substantial (20-30%) performance increase.