Message boards :
Number crunching :
I am getting a lot of gpu tasks with zero (0) expected processing times.
Message board moderation
Previous · 1 . . . 4 · 5 · 6 · 7 · 8 · 9 · Next
Author | Message |
---|---|
Bill Send message Joined: 30 Nov 05 Posts: 282 Credit: 6,916,194 RAC: 60 |
Well, it does not appear to be an upgrade for my APU. The Radeon system tray does not show the upgrade, nor does it when I look up my APU on AMD's website. It looks like it is available just for discrete graphics cards. Regardless, when I was driving home I realized you said this needs to be fixed on the server end, so AMD updating their drivers might be irrelevant? Or did I misunderstand you? Richard, as I was scrolling through this thread, I realized in this message you asked for the first few lines from the event log and I don't think I ever gave those to you. Do you want that still or is it moot at this point? Seti@home classic: 1,456 results, 1.613 years CPU time |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
Thanks, but we've actually moved on from that stage. The patched client seems to be showing what we need now - no further testing needed on that. Which is more than can be said for the servers - it became apparent on Wednesday that work is needed there too. I'm going to suggest to Eric that our 'SETI Beta' project might be a possible venue for testing, but I don't know if it's in the right condition to accept the preparatory code which has been written so far. Don't do anything yet, but if we do get a test started, would you be willing to attach your Ryzen to Beta briefly to see if it works? |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
Thanks, but we've actually moved on from that stage. The patched client seems to be showing what we need now - no further testing needed on that. My 2400G is back on the air. And I will make the requested client change. I can add Beta back if you want. Tom A proud member of the OFA (Old Farts Association). |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
Thanks. Please let us know whether - with the standard client - you see the crazy speed reported at startup. I've emailed Eric with the Beta suggestion, but he won't have received it yet because of the time zone difference. Hold off attaching to Beta until we know whether he can deploy something to test. |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
Thanks. Please let us know whether - with the standard client - you see the crazy speed reported at startup. I had to re-install Windows, then install the drivers on the CD-Rom and then re-install Seti. I am not seeing the really large Gflops number (yet). So either I am missing something. Or all my changes have "fixed" something. 2/15/2019 6:55:25 AM | | cc_config.xml not found - using defaults 2/15/2019 6:55:25 AM | | Starting BOINC client version 7.14.2 for windows_x86_64 2/15/2019 6:55:25 AM | | log flags: file_xfer, sched_ops, task 2/15/2019 6:55:25 AM | | Libraries: libcurl/7.47.1 OpenSSL/1.0.2g zlib/1.2.8 2/15/2019 6:55:25 AM | | Data directory: C:\ProgramData\BOINC 2/15/2019 6:55:25 AM | | Running under account tlgal 2/15/2019 6:55:25 AM | | OpenCL: AMD/ATI GPU 0: AMD Radeon(TM) RX Vega 11 Graphics (driver version 2482.7 (PAL,HSAIL), device version OpenCL 2.0 AMD-APP (2482.7), 6566MB, 6566MB available, 1761 GFLOPS peak) 2/15/2019 6:55:25 AM | | Host name: LYNNES-GIFT 2/15/2019 6:55:25 AM | | Processor: 8 AuthenticAMD AMD Ryzen 5 2400G with Radeon Vega Graphics [Family 23 Model 17 Stepping 0] 2/15/2019 6:55:25 AM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 htt pni ssse3 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 svm sse4a osvw skinit wdt tce topx page1gb rdtscp fsgsbase bmi1 smep 2/15/2019 6:55:25 AM | | OS: Microsoft Windows 10: Professional x64 Edition, (10.00.17134.00) 2/15/2019 6:55:25 AM | | Memory: 14.93 GB physical, 17.68 GB virtual 2/15/2019 6:55:25 AM | | Disk: 148.41 GB total, 117.02 GB free 2/15/2019 6:55:25 AM | | Local time is UTC -6 hours 2/15/2019 6:55:25 AM | | No WSL found. 2/15/2019 6:55:25 AM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 8671627; resource share 1000 2/15/2019 6:55:25 AM | SETI@home | General prefs: from SETI@home (last modified 13-Feb-2019 12:33:43) 2/15/2019 6:55:25 AM | SETI@home | Computer location: school 2/15/2019 6:55:25 AM | | General prefs: using separate prefs for school 2/15/2019 6:55:25 AM | | Reading preferences override file 2/15/2019 6:55:25 AM | | Preferences: 2/15/2019 6:55:25 AM | | max memory usage when active: 12996.41 MB 2/15/2019 6:55:25 AM | | max memory usage when idle: 14525.40 MB 2/15/2019 6:55:25 AM | | max disk usage: 100.00 GB 2/15/2019 6:55:25 AM | | max CPUs used: 7 2/15/2019 6:55:25 AM | | suspend work if non-BOINC CPU load exceeds 25% 2/15/2019 6:55:25 AM | | (to change preferences, visit a project web site or select Preferences in the Manager) 2/15/2019 6:55:25 AM | | Setting up project and slot directories 2/15/2019 6:55:25 AM | | Checking active tasks 2/15/2019 6:55:25 AM | | Setting up GUI RPC socket 2/15/2019 6:55:25 AM | | Checking presence of 281 project files A proud member of the OFA (Old Farts Association). |
Bill Send message Joined: 30 Nov 05 Posts: 282 Credit: 6,916,194 RAC: 60 |
I'd be happy to run the beta, but I'm not sure how to do that. Could point me in the right direction when the time is right? Seti@home classic: 1,456 results, 1.613 years CPU time |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
I had to re-install Windows, then install the drivers on the CD-Rom ...We think the problems arose in a more recent driver, downloaded from the internet. That would be a useful test: Bill was working fine when he first attached his Ryzen to SETI on 23 December, but he had his first failed task on 4 January. That sounds like driver 19.1 to me, so that's what I put in my bug report to AMD. If you could install something like that from the AMD website, that would be awesome. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
Could point me in the right direction when the time is right?Certainly - it only takes seconds. You just have to paste a url into the box provided. I'll tell you which url, and which box, nearer the time... |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
I had to re-install Windows, then install the drivers on the CD-Rom ...We think the problems arose in a more recent driver, downloaded from the internet. That would be a useful test: Bill was working fine when he first attached his Ryzen to SETI on 23 December, but he had his first failed task on 4 January. That sounds like driver 19.1 to me, so that's what I put in my bug report to AMD. Let me "update" to the latest. Since I have the cd-rom I can always "back date". :) Tom A proud member of the OFA (Old Farts Association). |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
If you do try to revert to the CD drivers after the test with the latest drivers, I would suggest using DDU to completely remove any vestiges of the troubled new drivers before reinstalling the good CD drivers. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
I had to re-install Windows, then install the drivers on the CD-Rom ...We think the problems arose in a more recent driver, downloaded from the internet. That would be a useful test: Bill was working fine when he first attached his Ryzen to SETI on 23 December, but he had his first failed task on 4 January. That sounds like driver 19.1 to me, so that's what I put in my bug report to AMD. I did a custom/clean install of the latest driver set for the all in one. 2/15/2019 12:16:54 PM | | cc_config.xml not found - using defaults 2/15/2019 12:17:02 PM | | Starting BOINC client version 7.14.2 for windows_x86_64 2/15/2019 12:17:02 PM | | log flags: file_xfer, sched_ops, task 2/15/2019 12:17:02 PM | | Libraries: libcurl/7.47.1 OpenSSL/1.0.2g zlib/1.2.8 2/15/2019 12:17:02 PM | | Data directory: C:\ProgramData\BOINC 2/15/2019 12:17:02 PM | | Running under account tlgal 2/15/2019 12:17:18 PM | | OpenCL: AMD/ATI GPU 0: AMD Radeon(TM) RX Vega 11 Graphics (driver version 2766.5 (PAL,HSAIL), device version OpenCL 2.0 AMD-APP (2766.5), 6566MB, 6566MB available, 1761 GFLOPS peak) 2/15/2019 12:17:18 PM | | Host name: LYNNES-GIFT 2/15/2019 12:17:18 PM | | Processor: 8 AuthenticAMD AMD Ryzen 5 2400G with Radeon Vega Graphics [Family 23 Model 17 Stepping 0] 2/15/2019 12:17:18 PM | | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 htt pni ssse3 fma cx16 sse4_1 sse4_2 movebe popcnt aes f16c rdrandsyscall nx lm avx avx2 svm sse4a osvw skinit wdt tce topx page1gb rdtscp fsgsbase bmi1 smep 2/15/2019 12:17:18 PM | | OS: Microsoft Windows 10: Professional x64 Edition, (10.00.17134.00) 2/15/2019 12:17:18 PM | | Memory: 14.93 GB physical, 17.68 GB virtual 2/15/2019 12:17:18 PM | | Disk: 148.41 GB total, 115.38 GB free 2/15/2019 12:17:18 PM | | Local time is UTC -6 hours 2/15/2019 12:17:18 PM | | No WSL found. 2/15/2019 12:17:19 PM | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 8671627; resource share 1000 2/15/2019 12:17:19 PM | SETI@home | General prefs: from SETI@home (last modified 13-Feb-2019 12:33:43) 2/15/2019 12:17:19 PM | SETI@home | Computer location: school 2/15/2019 12:17:19 PM | | General prefs: using separate prefs for school 2/15/2019 12:17:19 PM | | Reading preferences override file 2/15/2019 12:17:19 PM | | Preferences: 2/15/2019 12:17:19 PM | | max memory usage when active: 12996.41 MB 2/15/2019 12:17:19 PM | | max memory usage when idle: 14525.40 MB 2/15/2019 12:17:20 PM | | max disk usage: 100.00 GB 2/15/2019 12:17:20 PM | | max CPUs used: 7 2/15/2019 12:17:20 PM | | suspend work if non-BOINC CPU load exceeds 25% 2/15/2019 12:17:20 PM | | (to change preferences, visit a project web site or select Preferences in the Manager) 2/15/2019 12:17:20 PM | | Setting up project and slot directories 2/15/2019 12:17:20 PM | | Checking active tasks 2/15/2019 12:17:20 PM | | Setting up GUI RPC socket 2/15/2019 12:17:20 PM | | Checking presence of 281 project files A proud member of the OFA (Old Farts Association). |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
Agreed. That would be the Display Driver Uninstaller from https://www.wagnardsoft.com/, if you haven't come across it before: also distributed by Guru3D. |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
If you do try to revert to the CD drivers after the test with the latest drivers, I would suggest using DDU to completely remove any vestiges of the troubled new drivers before reinstalling the good CD drivers. I have been using the custom/clean install. If you have mentioned "DDU" before I failed to note it down. Found it, I think: https://www.guru3d.com/files-details/display-driver-uninstaller-download.html A proud member of the OFA (Old Farts Association). |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
OK, it looks as if your particular Ryzen doesn'r suffer from this bug. One of the other testers reported that it only showed up when he used a non-standard CPU speed (he downclocked it). Eric found host 8669611 while checking to see how widespread the problem is: user has been a SETI member since 2005, but has never posted on the message board. He attached a new Ryzen last Sunday (10 February), and had trashed 450 tasks by Wednesday. His CPU cores are working, so he probably hasn't even noticed. But the project - all projects - could do without that 'help', which is why I'm taking this so seriously. |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
OK, it looks as if your particular Ryzen doesn'r suffer from this bug. One of the other testers reported that it only showed up when he used a non-standard CPU speed (he downclocked it). Eric found host 8669611 while checking to see how widespread the problem is: user has been a SETI member since 2005, but has never posted on the message board. He attached a new Ryzen last Sunday (10 February), and had trashed 450 tasks by Wednesday. His CPU cores are working, so he probably hasn't even noticed. I will admit to turning off the "cool 'n quiet", the C states and setting the iGPU to forced so I can try to run a gtx 750Ti beside the iGPU like I did on the previous MB. But I haven't touched the cpu speed. This last re-boot I did set the iGPU to 1500 (it was stable on the last iGPU) and the iGPU voltage to 1.2. If it starts trashing tasks I will switch back to auto. --edit--- I just looked at the error listing on the website and I appear to have been throwing errors since yesterday. That was with the cpu/gpu on auto and the older drivers. The error is "exit time limit exceeded". Now what do I do? Shut down gpu processing? I want to leave it running for a little while because I have different drivers installed as well as the standard high power plan. Yesterday I was still running the "Amd balanced power plan" that even with "never stop" was pausing the system. --edit-- Tom A proud member of the OFA (Old Farts Association). |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
-edit---I'll go and look at them to see if it's the same bug we're trying to catch - that's certainly the I actually came here to report that Eric has just emailed me: I was planning on building updated server components today anyway... I'll see if I can get them deployed on beta before COB.His close of business is likely to be after my bedtime, but I'll keep an eye on it as long as I can. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
@ Tom M I see host 8671627 - Ryzen 5 2400G created 14 Feb 2019, 19:51:26 UTC. That sounds like the one. I also see 113 error tasks. The first one I spot-checked contains: Exit status 197 (0x000000C5) EXIT_TIME_LIMIT_EXCEEDED Run time 11 sec Device peak FLOPS 1,814,198.76 GFLOPS Max clock frequency: 42949672MhzThat's a full house of the problem symptoms we're tracking. So, please take a note of what settings you were running 24 hours ago, so that we can replicate them for Beta testing. BTW, that task was completed at 14 Feb 2019, 18:30:34 UTC, before the host was created. Trying to change too many things at once to get a host running quickly can cause problems like that. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
OK, another reply from Eric: Beta has been updated, and we're ready to start testing. First, please reduce you cache size/work request. We are looking for a problem which should only last for 10 tasks. If the solution works, you're cured for good, and that machine won't be useful for testing this type of problem ever again. Only if it fails can you help test the next attempted fix.... So, with a reduced cache setting, go to the tools menu in BOINC Manager. and click the first item, 'Add project...' Ignore all the information on existing projects - you won't find Beta there. Go straight to the last box, labelled 'Project URL:'. Paste in http://setiweb.ssl.berkeley.edu/beta/ Click 'Next >', and fill in your details. Please use the same email address and handle as you use elsewhere - it makes it easier to find you. The rest is self-explanatory. You'll get one task immediately for each of your devices. And then it'll ask for more work seven seconds later - you did set a small cache first, didn't you? Beta is different - no five minute pause to think about things. We're interested in the AMD GPU tasks only. Please note the initial runtime estimates, the actual runtimes, and anything else that catches your eye. Post it here, and we can pick over the entrails in the morning. Set 'No New Tasks' as soon as you want, and go back to your normal projects / cache settings. Thanks in advance. |
Richard Haselgrove Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874 |
Midnight passed, and my bed calls. Happy testing, and I'll look into any results you're able to gather in the morning. G'night. |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
@ Tom M It now occurs to me that the above creation date after the task was run has to be a result of the merge I did on the ids. A proud member of the OFA (Old Farts Association). |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.