F@H on ATI graphics cards....

Post by **Apoptosis** » Mon Oct 02, 2006 2:41 pm

not too shabby...

I'm working on Project: 2722 (Run 1, Clone 72, Gen 0)

[20:26:13] Completed 1
[20:34:09] Completed 2

It's only 220 points and has 2000 frames!!!

kenc51 · Post by **kenc51** » Mon Oct 02, 2006 2:44 pm

I'm using Riva Tuner to monitor the GFX card and I noticed it's running with 2D clocks!!!!!

Time to break out ATITool

Post by **dicecca112** » Mon Oct 02, 2006 2:46 pm

are you guys gonna get points for these WUs or what?

Post by **Apoptosis** » Mon Oct 02, 2006 2:53 pm

yes points count!

Post by **dicecca112** » Mon Oct 02, 2006 2:58 pm

hmm not on the normal point pages, I wonder what Protiens and there values they are assigning

Post by **Apoptosis** » Mon Oct 02, 2006 3:04 pm

why wouldn't they count on the normal pages?

I've seen people with:

p2711_GPUlambda
p2715_GPUlambda
p2717_p2714_GPUlambda
p2718_GPUlambda
p2722_GPUlambda

and they are all listed on the summary page.

Post by **dicecca112** » Mon Oct 02, 2006 3:11 pm

those so weren't there when I looked. Looks like 1 point or 220 pts. What about EMIII guys does it work with this?

Post by **Apoptosis** » Mon Oct 02, 2006 3:24 pm

[20:47:37] Completed 3
[20:55:20] Completed 4
[21:03:02] Completed 5
[21:10:44] Completed 6
[21:18:24] Completed 7
[21:22:07] mdrun_gpu returned 114
[21:22:07]
[21:22:07] Folding@home Core Shutdown: EARLY_UNIT_END
[21:22:09] CoreStatus = 72 (114)
[21:22:09] Sending work to server
[21:22:09] - Error: Could not get length of results file work/wuresults_01.dat
[21:22:09] - Error: Could not read unit 01 file. Removing from queue.
[21:22:09] - Preparing to get new work unit...
[21:22:09] + Attempting to get work packet
[21:22:09] - Connecting to assignment server
[21:22:09] - Successful: assigned to (171.65.103.159).
[21:22:09] + News From Folding@Home: Welcome to our GPU beta.
[21:22:10] Loaded queue successful

Well my first work unit ended in a failure.

Post by **dicecca112** » Mon Oct 02, 2006 3:28 pm

seem to be noticing at XS that OCed GPUs do not work well at all. You might wanna unoc it.

Post by **Apoptosis** » Mon Oct 02, 2006 3:31 pm

already did... it hasn't been the entire time I've been running the WU!

CoreStatus = 72 (114) seems to be a memory issue code. The EARLY_UNIT_END error is returned whenever a WU dies with a known error that can provide useful results. All this error code signifies is that the results can be returned even though the MD calculations failed. The points received are proportional to the work done, so at least I get a few points.

Stanford has this to say about my error:

We have yet to track down if this is an issue with the WU generation, or if we've hit a funny part of the code. Other projects are working okay...

Post by **dicecca112** » Mon Oct 02, 2006 3:38 pm

people have always been having issues install the DirectX from the FAQ, you sure its installed correctly. (Read I'm dying here not being able to install this on my 7900gt)

Post by **Apoptosis** » Mon Oct 02, 2006 5:31 pm

yes, this system has valid XP Pro on it and it's been updated...

Another one failed:

[20:47:37] Completed 3
[20:55:20] Completed 4
[21:03:02] Completed 5
[21:10:44] Completed 6
[21:18:24] Completed 7
[21:22:07] mdrun_gpu returned 114
[21:22:07]
[21:22:07] Folding@home Core Shutdown: EARLY_UNIT_END
[21:22:09] CoreStatus = 72 (114)
[21:22:09] Sending work to server
[21:22:09] - Error: Could not get length of results file work/wuresults_01.dat
[21:22:09] - Error: Could not read unit 01 file. Removing from queue.
[21:22:09] - Preparing to get new work unit...
[21:22:09] + Attempting to get work packet
[21:22:09] - Connecting to assignment server
[21:22:09] - Successful: assigned to (171.65.103.159).
[21:22:09] + News From Folding@Home: Welcome to our GPU beta.
[21:22:10] Loaded queue successfully.
[21:22:11] + Closed connections
[21:22:16]
[21:22:16] + Processing work unit
[21:22:16] Core required: FahCore_10.exe
[21:22:16] Core found.
[21:22:16] Working on Unit 02 [October 2 21:22:16]
[21:22:16] + Working ...
[21:22:16]
[21:22:16] *------------------------------*
[21:22:16] Folding@Home GPU Core - Beta
[21:22:16] Version 0.05 (Sat Sep 30 16:08:52 PDT 2006)
[21:22:16]
[21:22:16] Compiler : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 13.10.3077 for 80x86
[21:22:16] Build host: CYGWIN_NT-5.1 vishal-gpu 1.5.19(0.150/4/2) 2006-01-20 13:28 i686 Cygwin
[21:22:16] Preparing to commence simulation
[21:22:16] - Looking at optimizations...
[21:22:16] - Files status OK
[21:22:16] - Expanded 82082 -> 443705 (decompressed 540.5 percent)
[21:22:16]
[21:22:16] Project: 2722 (Run 1, Clone 72, Gen 0)
[21:22:16]
[21:22:16] Assembly optimizations on if available.
[21:22:16] Entering M.D.
[21:22:25] Completed 0
[21:22:25] Starting GUI Server
[21:30:06] Completed 1
[21:37:47] Completed 2
[21:45:27] Completed 3
[21:53:08] Completed 4
[22:00:52] Completed 5
[22:08:32] Completed 6
[22:16:11] Completed 7
[22:23:49] Completed 8
[22:31:28] Completed 9
[22:39:09] Completed 10
[22:46:51] Completed 11
[22:54:33] Completed 12
[23:02:16] Completed 13
[23:09:59] Completed 14
[23:16:23] mdrun_gpu returned 114
[23:16:23]
[23:16:23] Folding@home Core Shutdown: EARLY_UNIT_END
[23:16:26] CoreStatus = 72 (114)
[23:16:26] Sending work to server
[23:16:26] - Error: Could not get length of results file work/wuresults_02.dat
[23:16:26] - Error: Could not read unit 02 file. Removing from queue.
[23:16:26] - Preparing to get new work unit...

Post by **Apoptosis** » Mon Oct 02, 2006 5:40 pm

bad news... just sent some e-mails back and forth with Stanford and got this:

It looks like we have an issue with 6.10 and some of the work units. If you would be so kind to try 6.5, that would be great!

Doesn't look like things are going to well.

Post by **dicecca112** » Mon Oct 02, 2006 5:46 pm

yeah 6.10 doesn't fully support the x1950 right. Niksub over at OC forums, Ie one of the top 10 folders in the entire F@H project can't get it to work right either. After some digging he discovered that. And I don't believe the 6.5s support the 1950s

Post by **Apoptosis** » Mon Oct 02, 2006 6:44 pm

Nope... I have removed the 1950XTX and rolled back to the 1900XTX with 6.5 drivers...

I'm now working on project 2711

[00:23:45] Project: 2711 (Run 0, Clone 595, Gen 0)
[00:23:45]
[00:23:45] Assembly optimizations on if available.
[00:23:45] Entering M.D.
[00:23:54] Completed 0
[00:23:54] Starting GUI Server
[00:31:48] Completed 1
[00:39:41] Completed 2

kenc51 · Post by **kenc51** » Tue Oct 03, 2006 2:09 am

mine failed a few times last night.......the furthest it got was it completed 14 frames.......--> ~140mins worth.

I'll try the 6.5 drivers when I get home from work......

Post by **Apoptosis** » Tue Oct 03, 2006 7:24 am

Just woke up and my run of project 2711 seemed to stop working last night at 11:30pm which is when I left my PC... I guess the screen saver kicked on and stalled the client?

[00:52:49] Completed 3
[01:00:43] Completed 4
[01:08:38] Completed 5
[01:16:30] Completed 6
[01:24:22] Completed 7
[01:32:14] Completed 8
[01:40:07] Completed 9
[01:47:59] Completed 10
[01:55:51] Completed 11
[02:03:42] Completed 12
[02:11:36] Completed 13
[02:19:29] Completed 14
[02:27:23] Completed 15
[02:35:16] Completed 16
[02:43:19] Completed 17
[02:51:15] Completed 18
[02:59:18] Completed 19
[03:07:18] Completed 20
[03:15:11] Completed 21
[03:23:01] Completed 22
[03:30:52] Completed 23
[03:38:42] Completed 24
[03:46:33] Completed 25
[03:54:23] Completed 26
[04:02:13] Completed 27
[04:10:03] Completed 28
[04:17:54] Completed 29
[04:25:44] Completed 30
[04:33:35] Completed 31
[04:41:25] Completed 32
[04:49:15] Completed 33
[04:57:05] Completed 34
[05:04:56] Completed 35
[05:12:46] Completed 36
[05:20:36] Completed 37
[05:28:27] Completed 38
[05:36:17] Completed 39
[05:44:08] Completed 40
[05:51:58] Completed 41
[05:59:48] Completed 42
[06:07:38] Completed 43
[06:15:29] Completed 44
[06:23:19] Completed 45
[06:31:09] Completed 46
[06:38:59] Completed 47
[06:44:48] + Working...
[06:46:50] Completed 48
[06:54:40] Completed 49
[07:02:31] Completed 50
[07:10:21] Completed 51
[07:18:11] Completed 52
[07:26:01] Completed 53
[07:33:52] Completed 54
[07:41:42] Completed 55
[07:49:33] Completed 56
[07:57:23] Completed 57
[08:05:13] Completed 58
[08:13:04] Completed 59
[08:20:54] Completed 60
[08:28:44] Completed 61
[08:36:35] Completed 62
[08:44:25] Completed 63
[08:52:15] Completed 64
[09:00:05] Completed 65
[09:07:56] Completed 66
[09:15:46] Completed 67
[09:23:37] Completed 68
[09:31:26] Completed 69
[09:39:16] Completed 70
[09:47:07] Completed 71
[09:54:57] Completed 72
[10:02:48] Completed 73
[10:10:38] Completed 74
[10:18:28] Completed 75
[10:26:18] Completed 76
[10:34:09] Completed 77
[10:41:59] Completed 78
[10:49:50] Completed 79
[10:57:40] Completed 80
[11:05:30] Completed 81
[11:13:20] Completed 82
[11:21:11] Completed 83
[11:29:01] Completed 84
[11:36:52] Completed 85
[12:44:48] + Working...

kenc51 · Post by **kenc51** » Tue Oct 03, 2006 4:27 pm

p2711 Failed 11 times using the console version

p2725 Failed once just there after ~2hrs 15mins and now I've given up with the 6.10 drivers

Card wasn't overclocked during the above failures.

Post by **Apoptosis** » Tue Oct 03, 2006 5:27 pm

like I mentioned in an above posts 6.10 driver DO NOT work with F@H GPU Folding you HAVE TO roll back to 6.5 drivers.

Jeff_Grant · Post by **Jeff_Grant** » Tue Oct 10, 2006 12:08 pm

I've been running the console client on an x1600 Pro AGP card flawlessly with the 6.5 drivers. Had to manually install the x1900 driver in place of a x1600 driver but it still works great. The client gets to 85 like many others have reported, but keeps working and always finishes. No EUEs yet. Frame times on the x1600 are 21 min, but I don't think that's all that bad for this lower end card. I do OC it to 575 core 872 mem.