New GPU-Client with NV support: 6.12 beta

[Arnold Beckenbauer]

http://foldingforum.org/viewtopic.php?f=43&t=3186

We're happy to roll out the beta nVidia client, it's a nice beta, and so here we go...

First and before you do anything else, LOOK OVER THE FAQ. Lots of important things in there that you'll need to know if you do anything beyond run the installer and run it.

Drivers
You MUST use the 174.55 version of the drivers for now. You do not need the other files, just the drivers.

Known bugs:
It's a beta, and yes we meant that, not like Google but like software engineers - expect glitches and updates. The main thing is that the viewer will not work, as we're in the middle of changing how the live coordinates get to the viewer. Other then that, visit our nVidia beta Forum to see what's happening.

Download:
6.12 beta 6

[Jawed]

Yay.

I wonder if WUs can run on either ATI or NVidia or if each WU is created specifically for the type of GPU? I presume they'll run on either.

Jawed

[ShaidarHaran]

Can't wait to see what GT200 can pull off here. Same goes for RV770. Might put me back in contention for relevancy in the folding world once again, assuming my roomie goes with dual 4850s and I go single gtx 280 or 4870 x2/3/4 :D

Then again, I'm nearing the 3 million point barrier after < 4 years of folding, so I think I've done a fair share of work, although I feel irrelevant in the days of quad core machines pulling down 5k ppd, out-pacing my entire farm, which include(d) a PS3 folding 24x7, e8400 pulling down 2-3k ppd, and 30 or so odd machines out in the wild still crunching away. I'll get that PS3 up and running again now that I has intarwebs (just moved) but the PC will have to wait until I can get a different mobo (gremlins).

[Bludd]

Hey, this is pretty dope. I have a 8800 GTX with the 175.16 WHQL drivers for WinXP32. Works like a charm.

What is good performance in the client? I get 2000-3000 iterations/sec with 2100 being the most common one.

Edit: It's using quite a lot of CPU though, and my system is feeling unresponsive.

Edit2: I would have thought Geo would be all over this thread like a yuppie on coke.

[Geo]

Wheee. I may have to change my sig in about 15 mins. . .

Wish me luck. :)

[Arnold Beckenbauer]

Wheee. I may have to change my sig in about 15 mins. . .

Finally. :lol:

[Geo]

Bugger. No joy (yet).

I put the CUDA 174.55 on my Vista x64 machine with 8800GTX, but the FAH log file is telling me the thing keeps dying with:

[19:34:44] Folding@home Core Shutdown: UNSTABLE_MACHINE
[19:34:47] CoreStatus = 7A (122)

[Jawed]

64-bit Vista doesn't seem to work, check the FAQ.

Jawed

[Geo]

Arrrrgh. :cry:

[Jawed]

As an aside, I suppose this explains the dramatically higher PPD of NVidia GPUs:

http://foldingforum.org/viewtopic.php?p=30141#p30141

Thus far it has all been driver tuning and no tuning, yet, of the actual core in the public cores. Our concentration thus far has been on stability of the client and we are just starting a deep pass into optimization.

NVidia G92 GPUs appear to be generating 2-3x more PPD, around 5200 for a stock 8800GTS-512 it seems, versus 2000-2500 for RV670s.

Jawed

[Geo]

64-bit Vista doesn't seem to work, check the FAQ.

Jawed

Someone's worked out a driver hack to make GT200 drivers backwards compatible. Seems to work fine for me for folding anyway. No guarantee how'd it do in gaming.

http://foldingforum.org/viewtopic.php?f=43&t=3224

So. . . . drumroll please. . . .

Sig edited.

[Arnold Beckenbauer]

I've downloaded and installed 177.35 for my 8600GT (with crappy 1 GB DDR2!) yesterday with modified inf. I can fold, play and use h.264 accelleration.

[Geo]

Whee. GPU core running at 86C in a well ventilated case.

[Bludd]

I'm seeing 50% cpu utilization (1 of 2 cores) and if I scale it back using the control applet, I lose 70-90% of iter/sec

[Geo]

On an ATI card? There's been references to ATI requiring a lot more CPU for some reason at this stage.

I'm running about 4% cpu on my G80 client.

[Bludd]

No, 8800 GTX. WinXP32 SP3 FW 175.16.

http://img73.imageshack.us/img73/1637/fahcorecpubb4.th.png (http://img73.imageshack.us/img73/1637/fahcorecpubb4.png)

[Arnold Beckenbauer]

No, 8800 GTX. WinXP32 SP3 FW 175.16.

http://img73.imageshack.us/img73/1637/fahcorecpubb4.th.png (http://img73.imageshack.us/img73/1637/fahcorecpubb4.png)

Same here. (XP&177.35)
If someone gets problems with the forced core upgrade to 1.05: I had to uninstall the GPU client and to delete both folders, then it worked.

[Bludd]

It updated the core at first run, but that process went smoothly.

[Geo]

Anyone know if Quadros are going to be supported? I've got a Quadro FX 570 in my lappy I might like to give this a try on (Also Vista x64).

[CarstenS]

Everything "CUDA" runs just fine.

[BoardBonobo]

Does this work along side the SMP client?

[CarstenS]

Theoretically it should - creating around 10% CPU load on my C2D. But then, I'm running Vista x64 - and the SMP-client doesn't like that atm.

edit:
fixed - both are humming along nicely together. The SMP-client does not seem overly impaired by having the CPU feeding the GPU for the other client and still is supplying about 1.700 ppd.

[BoardBonobo]

Bugger. No joy (yet).

I put the CUDA 174.55 on my Vista x64 machine with 8800GTX, but the FAH log file is telling me the thing keeps dying with:

[19:34:44] Folding@home Core Shutdown: UNSTABLE_MACHINE
[19:34:47] CoreStatus = 7A (122)

OK. I'm getting the same error on a 4 core Vista Home Premium (32bit) box. Any ideas?

[CarstenS]

Tried the modified 177.39 drivers which were leaked a few days ago?

[Arnold Beckenbauer]

OK. I'm getting the same error on a 4 core Vista Home Premium (32bit) box. Any ideas?

There was a forced core upgrade to 1.05:
http://folding.typepad.com/news/2008/06/gpu2nvidia-auto-core-upgrade.html
GPU2/NVIDIA auto-core upgrade
We've forced a core upgrade to version 1.05 for GPU2/NVIDIA. We did this since there is important new code there. In particular, there's code that will better handle unstable or incorrectly configured machines. This core now generates an UNSTABLE_MACHINE error and the 6.12beta6 client will trap this error, leading to a 24 hour sleep for the client if 5 such UNSTABLE_MACHINE errors are reached.

If you are having problems running your client after this forced core upgrade, it most likely means that there's a problem with the client installation, most likely with startup aliases. If this is the case, please check out the FAQ and in particular note that the Start In and Target directories have to be different:

In Windows XP:
Target: "C:\Program Files\Folding@home\Folding@home-gpu\Folding@home.exe" -verbosity 9
(or whatever flags you use instead of -verbosity 9)
Start in: "C:\Documents and Settings\<your_windows_username>\Application Data\Folding@home-gpu\"

In Windows Vista:
Target: "C:\Program Files (x86)\Folding@home\Folding@home-gpu\Folding@home.exe" -verbosity 9
(or whatever flags you use instead of -verbosity 9)
Start in: "C:\Users\<your_windows_username>\AppData\Roaming\Folding@home-gpu\"

Check the forums:

http://foldingforum.org/viewforum.php?f=43&sid=b68878f4a886001a844dde1fd84dc55f

[Geo]

Does this work along side the SMP client?

Yep. :smile:

[Geo]

OK. I'm getting the same error on a 4 core Vista Home Premium (32bit) box. Any ideas?

Tried the modified 177.39 drivers which were leaked a few days ago?

Dunno about 177.39. I'm using 177.35 with a hacked .inf that guru3d did to support G80 and Quadros.

[BoardBonobo]

I upgraded to the modfied drivers and the UNSTABLE_MACHINE error has gone away. Now it's being replaced by Folding@home Core Shutdown: EARLY_UNIT_END. That and FILE_IO_ERROR. Is this normal for the beta?

And it's running fine along side the SMP client with no noticable sow downs.

Also why does there appear to be 2 instances of FahCore_11 running?

One last thing, will the folding client ever end up running on a PhysX card? Since it's basically a stream processor like a GPU?

[Geo]

Try running it as "run as administrator" if you're on Vista. That sounds like an authority problem.

[BoardBonobo]

Seems to be working fine now, no changes. 6 units done... faster than the PS3 but I guess you get less points for these?

[BoardBonobo]

I've just looked at the extreme overclocking site and realised that the drop in my output coincided with me playing MGS4. Freaky stats reading!!

[Chalnoth]

Try running it as "run as administrator" if you're on Vista. That sounds like an authority problem.
A lot of the time you can get rid of these sorts of issues by just adjusting the permissions in the program folder so that it's no longer "read only". Dunno if that'll work with folding at home, though.

[ShaidarHaran]

Great perf numbers for GPU client on G92 also with SMP client on dual-core 4GHz Penryn. Really juicy stuff here.

4GHz Penryn (E8400 @ 445x9)
4GB RAM @ 1GHz
8800 GT @ 685/1688/1940 on 177.35s - XP

4850 PPD for GPU client on its own. Combined with CPU SMP client (3600 GPU, 2300 CPU) 5900 PPD. Tweaked affinity and priorities brings this to 6450 total PPD (4550 GPU, 1900 CPU).

CPU SMP client all-out without GPU client running yields approximately 3000 PPD, so clearly a mix and match of the two clients yields the best results.

Temps don't really seem a concern. I have 6x 120mm case fans, a Xigmatek 964 HSF in push-pull config (dual 92mms full-out i.e. all thermal controls off), and the 88gt gets the TT duorb treatment with fans @ 70% via rivatuner.

[Arnold Beckenbauer]

Now larger proteins for NV GPUs, too: http://foldingforum.org/viewtopic.php?f=52&t=5452

[Spaceman-Spiff]

Now larger proteins for NV GPUs, too: http://foldingforum.org/viewtopic.php?f=52&t=5452

The NV GPUs slows down with larger proteins.

Some selected quotes:

---

by DanEnsign (http://foldingforum.org/memberlist.php?mode=viewprofile&u=16) on Mon Sep 08, 2008 6:56 pm

Hi,

I'm warning you, you will start to see a new series of projects coming out this morning. The proteins we're studying here are larger and in some sense more relevant to biology than anything we've run on GPU2 to date.

However, due to the larger size of the protein, you may see a PPD drop putting NV cards more in line with ATI production. This isn't due to different benchmarking or a different way to assign points, but rather we anticipate NV to be slightly less efficient with larger proteins than it has been with the smaller ones.

---

by John Naylor (http://foldingforum.org/memberlist.php?mode=viewprofile&u=227) on Mon Sep 08, 2008 9:11 pm CoreX5 wrote:Why if these projects are larger, are they worth less? Also why is the benchmark machine an ati 3850 when the same projects are not being run on nvidia as the ones being run on Ati clients? Just seeking some clarification as this could look like it was changed just to appease ati and their clients.
They are worth less because despite being bigger the benchmark machine processes them quicker than the old units (I think the simulations are shorter, just the proteins are bigger). The benchmark machine has always been a 3850 so nothing has changed.

And as an aside, the project staff have repeatedly said that nvidia's huge lead in the PPD stakes may change. It now has (I'm an nvidia folder too so I would be as miffed as anyone were it not for the fact that I don't give two hoots about points), so you can't say we weren't warned http://foldingforum.org/images/smilies/icon_smile.gif

---

by toTOW (http://foldingforum.org/memberlist.php?mode=viewprofile&u=76) on Mon Sep 08, 2008 10:32 pm

There is no point about calling it favoritism : we have always said that the first batch of WU (with ~550 atoms) was doing very well on the NV hardware and that might not be the case with future WU (see the "Preliminary results" in the PPD thread title).

It's just a matter of architecture : NV chose to use a few amount of complex shaders which are fast, but few. ATI chose simpler shader, which are a bit less efficient, but a lot.

On NV architecture, more atoms (and thus complexity and parallelism) means that there are no free SP to use, as they were already used hence the PPD drop.

On ATI hardware, there are still plenty SP available, so more complexity means better performance, as there were SP available.

We still see performance difference between two architectures because of SP efficiency, but both hardware tend to join their performances : NV slows a bit, and ATI gets a bit faster.


---

by slegrand (http://foldingforum.org/memberlist.php?mode=viewprofile&u=2833) on Mon Sep 08, 2008 10:50 pm

No worries, there are plenty of NV optimizations in the pipe as well, but I need to validate them before turning them loose and right now, I'm swamped with some other F@H-related work. As things stand, the NV algorithm is more efficient but the ATI algorithm is simpler. Hence around lambda-sized molecules, ATI gets a perf spike while NV is still warming up and bogged down by the housekeeping incurred for its more efficient algorithm. For Fip35-sized molecules, NV is underutilized on GTX280, but ATI is doubly so on 4870 and that's why the perf differential there is just crazy.

In the long run in this GPU generation, I *expect* SP for SP that an NV GPU will be about 2.5x faster than an ATI SP. This is of course then modified by SP count and processor clock. If Mike finds a structural optimization to his code, then I will revisit that ratio, but that's my prediction for now. And the game resets with the next hw generation anyway so it's an ongoing arms race.

[Florin]

Yup, all my Nvidia cores have received non-supervillins now and PPD is down across the board. Some rough preliminary PPD results from my FahMon;
GTX280 down from 6.8k to 5.7k
G92 GTS went from 4.8k to 3.4k
G80 GTS from 3.4k to 2.1k
G84 GT from 1.6k to 0.8k

But who could complain? The Nvidias have benefitted from excellent scores for a long time and those are still some great numbers. And after all it's all about the science :)

[Bludd]

Anyone else with this problem? Launching Process Explorer while this client is running crashes the system with an nvdisp4.dll error.

I am running the 182.05 betas

[ShaidarHaran]

Correct me if I'm wrong, but is there not a newer client available?

[Bludd]

Oh, yes. I downloaded the latest NV GPU client 6.23. It is with it I have the problem.