WARNING: This website is obsolete! Please follow this link to get to the new Albert@Home website!
Posts by zombie67 [MM] |
|
41)
Message boards :
Problems and Bug Reports :
Gpu tasks never start: status 'Scheduler wait' Windows 7 64 bit
(Message 111691)
Posted 14 Jan 2012 by  zombie67 [MM]
Post: We see this, too (on Mac OS). It's most likely related to a bug in the BOINC core client. A fix is already available and should be part of the next release (i.e. 7.0.8). Great news! 6.13/7.x was doing such funky stuff, I bought a replacement GPU because I thought that was the problem. Edit: Dang. No change. It still under-reports my GPU memory. Back to 6.12.35. |
|
42)
Message boards :
Problems and Bug Reports :
Running on ATI
(Message 111656)
Posted 5 Jan 2012 by zombie67 [MM]
Post: I am running 7.0.7, 12.1, XP64, C2Q, 5870, and the task is still consuming a about 40% of a CPU core. Edit: Some more info. When I reserve a core, GPU load is ~60%. When I do not reserve a core, GPU load drops to ~0%. Edit 2: Same behavior with 7.0.7, 12.1, Win7 64, i7 920, dual 5870s. |
|
43)
Message boards :
Problems and Bug Reports :
Gpu tasks never start: status 'Scheduler wait' Windows 7 64 bit
(Message 111588)
Posted 16 Dec 2011 by zombie67 [MM]
Post: Yeah, this new 6.13/7.x version of BOINC is persnickety when it comes to GPU memory. It is not about how much any more, is tis all about how much is "available" at BOINC start. |
|
44)
Message boards :
Problems and Bug Reports :
Catalysts 11.11, Windows 7 - 64bit.
(Message 111587)
Posted 16 Dec 2011 by zombie67 [MM]
Post: Huh. I have 11.11 and 2.5 installed, but I get the full CPU usage. No difference on win7 64. 100% CPU required per GPU (5870). I had 11.6 with SDK 2.5 installed before and I just installed the display driver and the updated SDK over the 11.6 ones (no uninstallation). Can you explain more, please? What updated SDK? I can't find anything higher than 2.5, which I already have. Which display driver? |
|
45)
Message boards :
Problems and Bug Reports :
Catalysts 11.11, Windows 7 - 64bit.
(Message 111526)
Posted 9 Dec 2011 by zombie67 [MM]
Post: Huh. I have 11.11 and 2.5 installed, but I get the full CPU usage. The only difference is I am on XP64. When I get back home next week, I will try it on a win7 machine. |
|
46)
Message boards :
Problems and Bug Reports :
BOINC 6.13.12 & ATI 11.8 Thru 11.11
(Message 111467)
Posted 1 Dec 2011 by zombie67 [MM]
Post: I take it back. Not fixed after all. |
|
47)
Message boards :
Problems and Bug Reports :
BOINC 6.13.12 & ATI 11.8 Thru 11.11
(Message 111465)
Posted 1 Dec 2011 by zombie67 [MM]
Post: Couldn't get the OpenCL Wu's to run on my 1 ATI 58xx Box with the v11.8 Drivers thru v11.11 Drivers, the Box would just freeze up, also the MOO Project wouldn't run with Client 6.13.12 installed, lost all running Wu's & Cached Wu's going back to Client 6.12.34. FYI, Moo! just updated their server SW, and the problem with the 6.13.xx has been resolved. |
|
48)
Message boards :
Problems and Bug Reports :
Running on ATI
(Message 111444)
Posted 30 Nov 2011 by zombie67 [MM]
Post: But running with all 4 cores, the time has shot way up. I don't have any examples to share yet. And at this rate, it looks like it will take about 12-13 hours to complete. So maybe tomorrow I will have an example to share. I think this means the task is CPU-starved. For the record, after about 6 hours, the whole system locked up and required a hard reboot. Upon reboot, the work for the ATI task was not lost. That's good. But I could not wait any longer (our team is joining another challenge today, which requires 6.12.xx or less). So I changed the CPU usage down to 50% to get it completed quickly. I won't have an example to show for a week or so. But I think we all know the behavior well enough now. With 11.11, the ATI tasks require a whole free core. |
|
49)
Message boards :
Problems and Bug Reports :
Running on ATI
(Message 111432)
Posted 29 Nov 2011 by zombie67 [MM]
Post: There is only one CPU in that machine. FWIW, run times with only 2 of 4 cores crunching are about the same as 3 of 4 cores. No noticeable difference. Below are some examples. But running with all 4 cores, the time has shot way up. I don't have any examples to share yet. And at this rate, it looks like it will take about 12-13 hours to complete. So maybe tomorrow I will have an example to share. I think this means the task is CPU-starved. With 50%: http://albert.phys.uwm.edu/result.php?resultid=45270 Run time 4,650.91 CPU time 4,331.19 http://albert.phys.uwm.edu/result.php?resultid=48300 Run time 5,166.38 CPU time 4,776.22 (FWIW, validated against CUDA) With 75%: http://albert.phys.uwm.edu/result.php?resultid=45059 Run time 5,165.22 CPU time 4,847.52 http://albert.phys.uwm.edu/result.php?resultid=45079 Run time 5,594.42 CPU time 4,749.09 Edit: One other note: For CUDA apps, the CPU time is about 25% of the run time. For these ATI tasks, it is about 95%. |
|
50)
Message boards :
Problems and Bug Reports :
Running on ATI
(Message 111427)
Posted 29 Nov 2011 by zombie67 [MM]
Post: About CPU usage and run times. Q9550 w/ 5870 , 11.11 I set BOINC to use only 3 of 4 cores. Running no GPU task, total CPU usage was ~80%. Run an albert ATI task and CPU usage jumped to 100%. Then set BOINC to use only 2 of 4 cores. Running no GPU task, total CPU usage was ~53%. Run an albert ATI task and CPU usage jumped to ~82%. Summary: This ATI app, on this hardware, needs (or at least uses) ~29% of the total CPU power, or slightly more than a full core. I am going to let it run a full task with BOINC using only two cores, then using three cores, and will report on run times. Edit: One more data point. BOINCmanager says this task is using 0.15 CPU + 1.00 GPU. I am not sure where that number comes from, but it clearly way off the mark. |
|
51)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111382)
Posted 21 Nov 2011 by zombie67 [MM]
Post: Ah. Well there you go. Good to know about the server-side stuff. |
|
52)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111380)
Posted 21 Nov 2011 by zombie67 [MM]
Post: Bernd wrote:BTW: In what sense is Albert@Home different from an alpha test project, in particular related to credit? Right. Albert is like RALPH, SETI Beta, and Amelegrid Testgrid. These are all dedicated test projects too. They will never progress to "production" status. The need for separate projects (any more) is not clear to me. A while back, they added the functionality in the BOINC server code to have opt-in on test apps/tasks for users. That allows the project to have test work available to only those that opt into it, in their project preferences. It seem to me that way would be less work for projects. But whatever, it works either way. |
|
53)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111374)
Posted 21 Nov 2011 by zombie67 [MM]
Post: If you (and the stat sites) can live with credit "however small it is", being granted unreliably and fluctuating, and won't complain whatever happens, I could as well leave the stats export on. After all, it's just a switch in the project configuration. That works for me. It is what happens at other test projects, so I would expect others would be fine with it too. Or at least they would be used to to it. I will be around regardless. |
|
54)
Message boards :
Problems and Bug Reports :
Bionc messages
(Message 111370)
Posted 21 Nov 2011 by zombie67 [MM]
Post:
6.13.xx (whatever is latest on the DL page). http://boinc.berkeley.edu/download_all.php |
|
55)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111366)
Posted 21 Nov 2011 by zombie67 [MM]
Post: Albert@Home stats are currently being exported, I intend to shut this down late this week. There will be a news item (including RSS) at least 24h in advance. Great news! Thanks! |
|
56)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111363)
Posted 21 Nov 2011 by zombie67 [MM]
Post: But I still like to understand what happened here, how & why people that are only interested in credit ever attached a host to this project. It's not either/or. At least not for me. I am here for *both* helping develop the project *and* getting credits for whatever happens to validate properly. Frankly, I am very interesting in getting the ATI app working properly. |
|
57)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111361)
Posted 21 Nov 2011 by zombie67 [MM]
Post: Seriously though : why was there any expectation of credit when exactly the reverse was advertised. Stats were being exported. That set the expectation that they would continue to be exported, unless or until there was an announcement of a change in policy. Simple as that. What non-sequitirs? Can you please be specific? You asked questions, so I thought I was being polite by answering. |
|
58)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111349)
Posted 21 Nov 2011 by zombie67 [MM]
Post: Look, do whatever you like. It's your project after all. I am only asking that stats be exported one last time (once the server is back up, and people have the opportunity to return the tasks), to cover the work in progress. |
|
59)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111348)
Posted 21 Nov 2011 by zombie67 [MM]
Post: Which stats sites do list Albert@Home, and why? Mundayweb & Free-DC are two that I know of. But I don't follow many stat sites. |
|
60)
Message boards :
Wish List :
Low credit for such a risky project
(Message 111341)
Posted 21 Nov 2011 by zombie67 [MM]
Post: I'm rather boggled by the queries here. Which part of "If you care about credit, this project is certainly not the right one for you." ( .... etc ) was unclear ? :-) The fact that one can lose crunching time and credits because of whatever reason (bad batch, bad app, etc), that part is always inherent in test projects. That risk is always part of the testing process, and people are used to that. The unclear (or rather the "surprising") part is where stats were being exported. This attracted a bunch of testers, and now they are being told the tasks in process will not be rewarded (assuming they validate). It would have been nice to be told in advance that this was going to change in advance. That would have given everyone time to empty their queues first, if they did not want to continue testing under the new policy change. I will ask again, Can the stats please be exported one last time (after the server has come back up and we have a change to return work in progress)? What is or is not done elsewhere ( +/- Mr Anderson ) is a non-question, and argument by plurality is a non-sequitir also. We aren't elsewhere, we're here. :-) My list was not an argument either for or against exporting stats. The project asked the question, so I answered the question for the project. Simple as that. Finally, I was never asking for more/less credits. I was (am) just asking for credits, whatever they may be, to continue to be exported. This was in another thread. But that other thread got locked, and discussion was redirected to this one. |
Previous 20 · Next 20
This material is based upon work supported by the National Science Foundation (NSF) under Grant PHY-0555655 and by the Max Planck Gesellschaft (MPG). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the investigators and do not necessarily reflect the views of the NSF or the MPG.
Copyright © 2024 Bruce Allen for the LIGO Scientific Collaboration