Message boards : Graphics cards (GPUs) : New beta application using cuda4.0 out for windows
Author | Message |
---|---|
This is for testing the new application on older cuda. | |
ID: 25333 | Rating: 0 | rate: / Reply Quote | |
Is it a problem to have only cuda4.2 for Linux? No problem for us. | |
ID: 25334 | Rating: 0 | rate: / Reply Quote | |
Hi, The latest drivers NVIDIA for Linux (Ubuntu) already support CUDA 4.2. Greetings. | |
ID: 25336 | Rating: 0 | rate: / Reply Quote | |
GTX560Ti CC2.1 XP 64bit driver 301.24, core 6.10.58: 3 tasks errored out at beginning | |
ID: 25337 | Rating: 0 | rate: / Reply Quote | |
This windows beta is giving major screen lag in XP 64. Driver 301.42, 7.0.25, GTX580. Have suspended while I still can. One completed OK & one failed after 20secs. | |
ID: 25338 | Rating: 0 | rate: / Reply Quote | |
This is for testing the new application on older cuda. The priority needs to be set to below normal, because these units make the computer slow and sluggish. The cuda 4.0 units are about 50+ seconds slower than the cuba 4.2 units on my computers, so why bother with cuba 4.0 all together, and have everyone update their drivers? | |
ID: 25340 | Rating: 0 | rate: / Reply Quote | |
On my 680 I'm receiving both CUDA 4 and 4.2. Mainly 4.2, and as expected the 4.0 error out at 0.00 | |
ID: 25341 | Rating: 0 | rate: / Reply Quote | |
Assuming the cuda40 and 42 beta work units are the same size the cuda42 seems to be the fastest on cpu and run-time on my GTX570. Link to one host here | |
ID: 25342 | Rating: 0 | rate: / Reply Quote | |
A fix coming up tomorrow. Now you should be getting always the same application. | |
ID: 25343 | Rating: 0 | rate: / Reply Quote | |
The cuda 4.0 units are about 50+ seconds slower than the cuba 4.2 units on my computers, so why bother with cuba 4.0 all together, and have everyone update their drivers? The project is trying to be nice to crunchers. For some updating the driver is a nightmare or not [easily] possible due to various reasons. MrS ____________ Scanning for our furry friends since Jan 2002 | |
ID: 25347 | Rating: 0 | rate: / Reply Quote | |
From discussions with the Seti optimisers (well one of them anyway), he said that the older GTX2xx cards are worse off running cuda42 apps. It would seem cuda42 is particularly optimised for Fermi and Kepler chips. The old cards can still run them, they are slower than a cuda32 app on the same card. I don't think they ever tried a cuda40 app, but they did try cuda41 which also was worse than cuda32 on older cards. | |
ID: 25348 | Rating: 0 | rate: / Reply Quote | |
"nenym" wrote: GTX560Ti CC2.1 XP 64bit driver 301.24, core 6.10.58: 3 tasks errored out at beginningThe same one beta CUDA 42 task with 301.42 driver. | |
ID: 25365 | Rating: 0 | rate: / Reply Quote | |
The new app does not compile on anything older than cuda4.0. From discussions with the Seti optimisers (well one of them anyway), he said that the older GTX2xx cards are worse off running cuda42 apps. It would seem cuda42 is particularly optimised for Fermi and Kepler chips. The old cards can still run them, they are slower than a cuda32 app on the same card. I don't think they ever tried a cuda40 app, but they did try cuda41 which also was worse than cuda32 on older cards. | |
ID: 25366 | Rating: 0 | rate: / Reply Quote | |
The new app does not compile on anything older than cuda4.0. If cuda40 is the minimum version you can do, can you supply it based on the compute capability of the card? ____________ BOINC blog | |
ID: 25376 | Rating: 0 | rate: / Reply Quote | |
New cuda4.0 app is out now for windows only. | |
ID: 25380 | Rating: 0 | rate: / Reply Quote | |
The first task (cuda40) took 2 minutes & all was fine. The other tasks (cuda42) are causing the screen to become unresponsive. Tasks finishing without error. XP64,GTX580, 7.0.28, 301.42. | |
ID: 25384 | Rating: 0 | rate: / Reply Quote | |
Welcome to the 2D clock bug | |
ID: 25388 | Rating: 0 | rate: / Reply Quote | |
Yeah. Just rebooted & back to normal. | |
ID: 25390 | Rating: 0 | rate: / Reply Quote | |
Is yours a factory overclocked card? | |
ID: 25391 | Rating: 0 | rate: / Reply Quote | |
I flashed the bios myself a year ago & never had any issues with it. I think the driver just crashed rather than down clocked. That slow cuda40 task was straight after a cuda42 one which was causing lag. | |
ID: 25392 | Rating: 0 | rate: / Reply Quote | |
GDF: | |
ID: 25393 | Rating: 0 | rate: / Reply Quote | |
I removed the cuda4.2 application. There should only be a 4.0 one with no lag problems. | |
ID: 25395 | Rating: 0 | rate: / Reply Quote | |
cuda40 tasks seem to be all ok with no lag. However, should they be using 100% cpu thread with no swan_sync? | |
ID: 25396 | Rating: 0 | rate: / Reply Quote | |
no. it should not. | |
ID: 25397 | Rating: 0 | rate: / Reply Quote | |
I removed the cuda4.2 application. There should only be a 4.0 one with no lag problems. I am still getting the cuda4.2, with the priority set at high, which causing the computer to be sluggish. The cuda4.0 have a below normal priority, with no sluggishness. | |
ID: 25399 | Rating: 0 | rate: / Reply Quote | |
can you reset the project and try again? | |
ID: 25401 | Rating: 0 | rate: / Reply Quote | |
Is it a problem to have only cuda4.2 for Linux? The Ubuntu 12.04 repository version of the NVidia driver is 295.40. This is CUDA 4.2 capable. The more recent 295.53 Certified version can also be downloaded from NVidia. It might be worth noting however that for people with non-fermi/kepler cards (probably not very many) a CUDA 4.2 app would perform less efficiently, and for those with older drivers and Fermi or Kepler cards they would not be able to contribute without updating drivers. Is a project reset now required for the new/chosen app to propagate, rather than a standard project update? On swan_sync. Changes to this require a restart to take effect (just in case you didn't). Sounds like 'real-time' priority though. BTW, is there anyone crunching for GPUGrid that still uses Win98? http://www.gpugrid.net/apps.php Platform - Version - Installation time Microsoft Windows (98 or later) - running on an Intel x86-compatible CPU - 31 May 2012 | 17:48:34 UTC ____________ FAQ's HOW TO: - Opt out of Beta Tests - Ask for Help | |
ID: 25402 | Rating: 0 | rate: / Reply Quote | |
can you reset the project and try again? How would that work, my machine just asking for work? Your server is sending it out. | |
ID: 25403 | Rating: 0 | rate: / Reply Quote | |
can you reset the project and try again? OK, I reset the project on one of my machines, after that it downloaded 2 cuda4.2 units. The priority is still set high. 5/31/2012 7:42:15 PM | GPUGRID | Scheduler request completed 5/31/2012 7:42:45 PM | GPUGRID | Resetting project 5/31/2012 7:42:52 PM | GPUGRID | work fetch resumed by user 5/31/2012 7:42:55 PM | GPUGRID | update requested by user 5/31/2012 7:43:00 PM | GPUGRID | Sending scheduler request: Requested by user. 5/31/2012 7:43:00 PM | GPUGRID | Requesting new tasks for NVIDIA 5/31/2012 7:43:05 PM | GPUGRID | Scheduler request completed: got 2 new tasks 5/31/2012 7:43:07 PM | GPUGRID | Started download of acemd.2517.win 5/31/2012 7:43:07 PM | GPUGRID | Started download of cudart32_42_9.dll 5/31/2012 7:43:20 PM | GPUGRID | Finished download of cudart32_42_9.dll 5/31/2012 7:43:20 PM | GPUGRID | Started download of cufft32_42_9.dll 5/31/2012 7:43:34 PM | GPUGRID | Finished download of acemd.2517.win 5/31/2012 7:43:34 PM | GPUGRID | Started download of tcl85.dll 5/31/2012 7:43:40 PM | GPUGRID | Sending scheduler request: To fetch work. 5/31/2012 7:43:40 PM | GPUGRID | Requesting new tasks for NVIDIA 5/31/2012 7:43:42 PM | GPUGRID | Finished download of tcl85.dll 5/31/2012 7:43:42 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-LICENSE 5/31/2012 7:43:43 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-LICENSE 5/31/2012 7:43:43 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-COPYRIGHT 5/31/2012 7:43:43 PM | GPUGRID | Scheduler request completed: got 0 new tasks 5/31/2012 7:43:43 PM | GPUGRID | No tasks sent 5/31/2012 7:43:43 PM | GPUGRID | This computer has reached a limit on tasks in progress 5/31/2012 7:43:45 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-COPYRIGHT 5/31/2012 7:43:45 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-coor_file 5/31/2012 7:43:54 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-coor_file 5/31/2012 7:43:54 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-vel_file 5/31/2012 7:44:03 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-vel_file 5/31/2012 7:44:03 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-idx_file 5/31/2012 7:44:04 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-idx_file 5/31/2012 7:44:04 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-pdb_file 5/31/2012 7:44:25 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-pdb_file 5/31/2012 7:44:25 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-psf_file 5/31/2012 7:44:54 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-psf_file 5/31/2012 7:44:54 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-par_file 5/31/2012 7:44:59 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-par_file 5/31/2012 7:44:59 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-conf_file_enc 5/31/2012 7:45:00 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-conf_file_enc 5/31/2012 7:45:00 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-metainp_file 5/31/2012 7:45:01 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-metainp_file 5/31/2012 7:45:01 PM | GPUGRID | Started download of 291-GIANNI_TEST6-0-hills_file 5/31/2012 7:45:02 PM | GPUGRID | Finished download of 291-GIANNI_TEST6-0-hills_file 5/31/2012 7:45:02 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-LICENSE 5/31/2012 7:45:03 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-LICENSE 5/31/2012 7:45:03 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-COPYRIGHT 5/31/2012 7:45:04 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-COPYRIGHT 5/31/2012 7:45:04 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-311-GIANNI_TEST6-0-5-RND7130_1 5/31/2012 7:45:09 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-311-GIANNI_TEST6-0-5-RND7130_1 5/31/2012 7:45:09 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-311-GIANNI_TEST6-0-5-RND7130_2 5/31/2012 7:45:18 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-311-GIANNI_TEST6-0-5-RND7130_2 5/31/2012 7:45:18 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-311-GIANNI_TEST6-0-5-RND7130_3 5/31/2012 7:45:21 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-311-GIANNI_TEST6-0-5-RND7130_3 5/31/2012 7:45:21 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-pdb_file 5/31/2012 7:45:38 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-pdb_file 5/31/2012 7:45:38 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-psf_file 5/31/2012 7:46:14 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-psf_file 5/31/2012 7:46:14 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-par_file 5/31/2012 7:46:17 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-par_file 5/31/2012 7:46:17 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-conf_file_enc 5/31/2012 7:46:18 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-conf_file_enc 5/31/2012 7:46:18 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-metainp_file 5/31/2012 7:46:19 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-metainp_file 5/31/2012 7:46:19 PM | GPUGRID | Started download of 311-GIANNI_TEST6-1-311-GIANNI_TEST6-0-5-RND7130_7 5/31/2012 7:46:20 PM | GPUGRID | Finished download of 311-GIANNI_TEST6-1-311-GIANNI_TEST6-0-5-RND7130_7 5/31/2012 7:46:20 PM | GPUGRID | Started download of logogpugrid.png 5/31/2012 7:46:21 PM | GPUGRID | Finished download of logogpugrid.png 5/31/2012 7:46:21 PM | GPUGRID | Started download of project_1.png 5/31/2012 7:46:22 PM | GPUGRID | Finished download of project_1.png 5/31/2012 7:46:22 PM | GPUGRID | Started download of project_2.png 5/31/2012 7:46:23 PM | GPUGRID | Finished download of project_2.png 5/31/2012 7:46:23 PM | GPUGRID | Started download of project_3.png 5/31/2012 7:46:24 PM | GPUGRID | Finished download of project_3.png 5/31/2012 7:46:48 PM | GPUGRID | Finished download of cufft32_42_9.dll 5/31/2012 7:46:49 PM | GPUGRID | Starting task 291-GIANNI_TEST6-0-5-RND7418_2 using acemdbeta version 643 (cuda42) in slot 1 | |
ID: 25407 | Rating: 0 | rate: / Reply Quote | |
------------------------------------------------- | |
ID: 25413 | Rating: 0 | rate: / Reply Quote | |
GTX560Ti CC2.1, XP64bit, driver 301.42, Swan_Sync NOT set. | |
ID: 25415 | Rating: 0 | rate: / Reply Quote | |
You need to upgrade Boinc to at least 7.0.25. I'm using beta 7.0.28 & it's OK for me. | |
ID: 25416 | Rating: 0 | rate: / Reply Quote | |
Is upgrading to 7.0.25 of the Boinc core only sufficient? I hate the 7.x manager. | |
ID: 25417 | Rating: 0 | rate: / Reply Quote | |
"Up"grade to 7.0.25 did not help. | |
ID: 25419 | Rating: 0 | rate: / Reply Quote | |
Is upgrading to 7.0.25 of the Boinc core only sufficient? I hate the 7.x manager. I agree.... I much prefer 6.10.60 | |
ID: 25420 | Rating: 0 | rate: / Reply Quote | |
Methinks it's a larger problem than what NV diriver is installed. I'm still hanging in 3.1 waiting for the fix... NV570 X2 (with 7.0.25 client) | |
ID: 25426 | Rating: 0 | rate: / Reply Quote | |
Does that mean we must upgrade boinc too not only to graphics 301.x driver | |
ID: 25428 | Rating: 0 | rate: / Reply Quote | |
You don't need to upgrade to the latest version of Boinc. | |
ID: 25432 | Rating: 0 | rate: / Reply Quote | |
You don't need to upgrade to the latest version of Boinc. Nice one!!! | |
ID: 25437 | Rating: 0 | rate: / Reply Quote | |
Another application for cuda4.0 is out only windows. | |
ID: 25438 | Rating: 0 | rate: / Reply Quote | |
FYI - The server is sending these cuda 4.0 WUs to my GTX670 (Win7x64). I'll let it fail and try a reset but it didn't help last time. | |
ID: 25440 | Rating: 0 | rate: / Reply Quote | |
Three tasks received. All failed after a few seconds.XP64,GTX580, 7.0.28, 301.42. | |
ID: 25441 | Rating: 0 | rate: / Reply Quote | |
Three tasks received. All failed after a few seconds.XP64,GTX580, 7.0.28, 301.42. Hi everybody. While I was trying to do my usual 2 longrun this weekend, I noticed them failed almost immediately. And also 6 more failed with various errors :( There are all (2 still "running" are just because I resetted before they sent, but crashed too) http://www.gpugrid.net/results.php?userid=86407 :( Bye, R! | |
ID: 25444 | Rating: 0 | rate: / Reply Quote | |
All units are failing after a few seconds: | |
ID: 25445 | Rating: 0 | rate: / Reply Quote | |
Tonight I have been getting a bunch of beta 6.45 cuda40 work units. All failed. Most with "energies have become nan" around 19% through. I have suspended work fetch so others can try and process them. | |
ID: 25448 | Rating: 0 | rate: / Reply Quote | |
I'm getting failures too. I think these may be around the checkpoint time? | |
ID: 25452 | Rating: 0 | rate: / Reply Quote | |
Is it just the Windows cuda40 application that's failing, or the cuda42 one as well? | |
ID: 25453 | Rating: 0 | rate: / Reply Quote | |
GTX670, Win7 x64, BOINC 7.0.25 x64, GeForce 301.42 | |
ID: 25454 | Rating: 0 | rate: / Reply Quote | |
Same here - all I am receiving is the Cuda 4.0 work which fails after a few seconds so clearly some problem back server side as it should not be sending us 4.0 work in first place. | |
ID: 25455 | Rating: 0 | rate: / Reply Quote | |
Is it just the Windows cuda40 application that's failing, or the cuda42 one as well? I am not getting cuda42. I only seem to be getting cuda40 work. Suspending work fetch again as no point failing all of them. The last cuda42 I got was a v6.43 (a couple of days ago) and that worked fine. [edit] I will add that deselecting Beta tasks in preferences, getting client to update when set to NNT and then asking for work still gives beta work. Looks like I won't be crunching GPUgrid for a few days until the beta queue gets drained [/edit] ____________ BOINC blog | |
ID: 25462 | Rating: 0 | rate: / Reply Quote | |
To stop getting these Beta's you have to say No to the Beta app: | |
ID: 25473 | Rating: 0 | rate: / Reply Quote | |
MJH, Perhaps Gianni removed the CUDA 4.2 app to just test 4.0 apps? Looks like they'll have a busy monday :D MrS ____________ Scanning for our furry friends since Jan 2002 | |
ID: 25475 | Rating: 0 | rate: / Reply Quote | |
Yeah the beta apps are failing for me too when the test units last week were doing fine. | |
ID: 25477 | Rating: 0 | rate: / Reply Quote | |
It would be useful to be able to see the app version names in the Server Status page. | |
ID: 25481 | Rating: 0 | rate: / Reply Quote | |
GTX560Ti CC2.1, driver 301.42, XP 64bit, core 7.0.25, BM 6.10.58, Swan_Sync NOT set. | |
ID: 25483 | Rating: 0 | rate: / Reply Quote | |
Win XP pro 32bit sp3 , BM 7.0.25, Nv driver 301.42 | |
ID: 25488 | Rating: 0 | rate: / Reply Quote | |
Win XP pro 32bit sp3 , BM 7.0.25, Nv driver 301.42 GTX 470 also failed with another NV driver 285.85 on ACEMD beta v6.45(cuda40) ____________ | |
ID: 25525 | Rating: 0 | rate: / Reply Quote | |
Message boards : Graphics cards (GPUs) : New beta application using cuda4.0 out for windows