Folding@Home Benchmark 1.2.0 (GUI)

[proteneer]

Napolean,

PS make sure to redownload the .zip file - I posted an update earlier (about 3 hours ago)

what error does it throw? the CUDA platform should work.

Also, try running both commands:

FAHBench.exe -deviceID 0 -platform OpenCL -precision single
FAHBench.exe -deviceID 0 -platform CUDA -precision single

[k1wi]

Having installed 08 from your link and searching my entire computer, it would appear that cl.exe still only exists in the VS2012 folder? Never mind, Looks like i installed visual basic instead, having another go.

[proteneer]

hm.. did you download vcsetup.exe? (NOT vcssetup.exe) It should be under C:\Program Files (x86)\Microsoft Visual Studio 9.0\VC\bin

I shoulda made that more clear - modified the page to reflect it.

[k1wi]

Managed to get a successful run after installing vcsetup.

570 @ 849MHz:
CUDA:
Explicit: 25.3849
Implicit: 96.9697

Haven't managed to pull a OpenCL result for my GPU.

[proteneer]

Managed to get a successful run after installing vcsetup.

570 @ 849MHz:
CUDA:
Explicit: 25.3849
Implicit: 96.9697

Haven't managed to pull a OpenCL result for my GPU.

Thanks! Those are pretty good numbers. What error do you get with OpenCL?

[k1wi]

570 @952 GPU, 2102 memory (the max I fold at)
Explicit: 28.3186
Implicit: 106.667

I just wanted to beat that 660Ti

[proteneer]

570 @952 GPU, 2102 memory (the max I fold at)
Explicit: 28.3186
Implicit: 106.667

I just wanted to beat that 660Ti

You mean the $3000 Telsa K20 card? Congrats Do you overclock?

[proteneer]

Yeah, the stock 570 GPU clock is 732Mhz, so first run was 849 (15% OC) and second was 952 (30% OC)

At stock it gets:
22.3066
85.4938

Just for giggles, my i7 920 @4Ghz managed
OpenCL:
1.79233 ns/day
3.36443 ns/day

The bench app does utilize an entire CPU core during long periods when benching the GPU.

LOL I actually had no idea the OpenCL client works on Intel CPUs. Last time we checked it was quite buggy but I'm surprised you got it to fold.

[Napoleon]

I'm still out of luck getting it to run, but at least my environment should be more or less OK now. Command Prompt was started with "Run as administrator". I ran vcvarsall.bat to set up the environment variables. So what might still be going wrong here? At least cl.exe & nvcc.exe seem to be launching as they should now:

Code: Select all

Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation.  All rights reserved.

C:\Windows\system32>cd \temp\FAHBench_0_1

C:\Temp\FAHBench_0_1>pushd "\Program Files (x86)\Microsoft Visual Studio 9.0"\vc

C:\Program Files (x86)\Microsoft Visual Studio 9.0\VC>vcvarsall.bat
Setting environment for using Microsoft Visual Studio 2008 x86 tools.

C:\Program Files (x86)\Microsoft Visual Studio 9.0\VC>popd

C:\Temp\FAHBench_0_1>FAHBench.exe -deviceID 0 -platform CUDA -precision single

                                          O              O
   P R O T E N E E R     C--N              \              \               N
                         |                  C              C=O           / \-C
                         C                 /               |          N-C     \
  .C-C                 C/                  C               C           |      C
 /    \          O     |                   |               /           N      |
C     C          |     |           O       C              C                 /-C
 \_N_/ \   N    _C_    C           |      /         O    /                 C
        C-/ \_C/   \N-/ \    N   /-C-\   C          |    |           O    /
        |     |           C-/ \C/     N-/ \_   N\  /C\  -C      N    |    |
        O     |           |    |            \C/  C/   N/  \_C__/ \   C-\  C
              C           O    |             |   |          |     C-/   N/ \-C
               \_C             C             O   |          O     |          |
                  \             \-O              C                C          O
                  |                               \                \
                  C    N         Folding@Home      C--N             C
                   \   |      Benchmark  (Beta)    |                |
                    N--C                           O                |
                        \        Yutong Zhao                       C=O
                         N    proteneer@gmail.com                 /
                                                                 O

               for official stats, please visit www.fahbench.com

Invalid Device (please set valid deviceID by using --display-devices)

C:\Temp\FAHBench_0_1>cl
Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 15.00.21022.08 for 80x86
Copyright (C) Microsoft Corporation.  All rights reserved.

usage: cl [ option... ] filename... [ /link linkoption... ]

C:\Temp\FAHBench_0_1>nvcc
nvcc : fatal error : No input files specified; use option --help for more information

C:\Temp\FAHBench_0_1>

[k1wi]

Napoleon - use -deviceId rather than -deviceID

[proteneer]

edit - what kiwi said =)

[proteneer]

Well, at least FAHBench.exe -deviceId 0 -platform OpenCL -precision single got something done:

Code: Select all

Explicit:
Checking for accuracy...done
3.28365 ns/day
Implicit:
Checking for accuracy...done
18.4848 ns/day

It took a while... but I suppose that was to be expected on a lowly GT430, right? Oh well, at least it beat the i7 920 OCL result.

FAHBench.exe is taking up one CPU core entirely (25% utilization on a 2C/4T CPU). I presume that's normal, too?

Unfortunately I wasn't so lucky with FAHBench.exe -deviceId 0 -platform CUDA -precision single

Code: Select all

Explicit:
Error launching CUDA compiler: -1
nvcc : fatal error : Value 'compute_21' is not defined for option 'gpu-architecture'

Congrats you got it working!

The compute_21 error is a bug on our end, sorry about that.

[P5-133XL]

I'm getting a new error. So what did I do wrong?

Error launching CUDA compiler: -1
nvcc : Fatal error : value 'compute_21' is not defined for option 'gpu-architecture'

[Napoleon]

Nothing I guess, viewtopic.php?f=16&t=23434&start=15#p233690.

[proteneer]

Ah yes, we've never tested much of OpenMM on IONs - it doesn't seem to support atomics.