Posts by Graham Thomas


log in
1) Message boards : Number crunching : Can I prevent longer 7.03 WUs being downloaded? (GPU problem) (Message 1653136)
Posted 15 Mar 2015 by Profile Graham ThomasProject donor
Thanks for all that extra info. Very useful.

I've now detached from and reattached to the project. I noticed that the apps were downloaded again, so I guess that the recompilations have been done.

I tried leaving some of the longer 7.03 units in the list to see if they would complete. The first one that was processed did stall, but it got further along (ca 60%) than was the case previously. Now I'll try leaving just the 'shorter longer' units (if that makes sense - those scheduled to take, say, 2 hrs 15 mins to finish rather than the 3 hrs or more scheduled for the longest ones) and see what happens.

I'll then try adding the line suggested above to the two files.

(Oh, and I also tried letting Device Manager update my display driver, just in case my AMD Catalyst suite wasn't up-to-date. Big mistake. It 'updated' to a 2011 driver and totally knocked out the GPU. Thank goodness for the roll-back feature.)

Admiral Gloval: it looks like your machine is processing the 7.03 units about twice as fast as mine. How good that it depends of course on how capable your processor is supposed to be (there are comparison tables you can search for), but from here it looks OK.
2) Message boards : Number crunching : Can I prevent longer 7.03 WUs being downloaded? (GPU problem) (Message 1652464)
Posted 13 Mar 2015 by Profile Graham ThomasProject donor
OK, now I'm a bit confused.

Some of the results reported above (especially the Astropulse 6 result) are from several months ago, before I updated my ATI drivers. They were certainly old then, and I updated after I saw that Astropulse 7 units were exiting (or stalling - can't remember - as soon as they started.

You can see what I'm running now from my post at the top of this thread. I've just checked my AMD Catalyst Control Center and the software update checker tells me I'm up-to-date (I'm running the standard version, not the beta.) The Catalyst software suite version is given as 14.501.1003.etc

I can see a CAL version from my seti@home event log. It's 1.4.1848, which is higher than the minimum given in posts above. My OpenCL version is 1.2 and as far as I know it just comes with the Catalyst suite. I can't see any mention of SDK.

I can certainly try detaching from the project and reattaching. But what would that change?
3) Message boards : Number crunching : Can I prevent longer 7.03 WUs being downloaded? (GPU problem) (Message 1652028)
Posted 12 Mar 2015 by Profile Graham ThomasProject donor
Thanks, Mike.

However, I can't see a file called

mb_cmdline_win_x86_SSE_OpenCL_ATi_HD5.txt

in my

C:\ProgramData\BOINC\projects\setiathome.berkeley.edu

folder (which I hope is what you meant by 'projects' folder).

I do, though, have two (empty) files there called:

mb_cmdline_7.03_opencl_ati_sah.txt
mb_cmdline_7.03_opencl_ati5_sah.txt

Would it be useful to add the -period_iterations_num line to either or both of these?
4) Message boards : Number crunching : Can I prevent longer 7.03 WUs being downloaded? (GPU problem) (Message 1651592)
Posted 11 Mar 2015 by Profile Graham ThomasProject donor
I haven't seen any Astropulse units for a long time now, but I do get lots of seti@home v7 7.03 WUs downloaded for my GPU. This in itself isn't a problem, but units scheduled to take a longer time to complete end up stalling. They reach a certain percentage of their worktime (usually around 20%), then stick at that percentage and amount of time completed, while incrementing the time to completion. Also, at the point when the WU stalls, my screen display goes blank and then recovers.

These WUs are typically scheduled to take 2-3 hours on my machine, whereas shorter WUs that are scheduled to complete in just over an hour run without problems. I've been told (in response to an earlier post here) that there are some combinations of GPU hardware and drivers that just don't handle the longer units. (Details of my setup are below.)

I've got used to aborting the longer WUs before they run, but it's tedious. Is there a way to make sure they don't download in the first place? From the spec of the cc_config.xml file, it looks like it might be possible to stop *all* v7.03 WUs from downloading (via the exclude_gpu option, though I'm no expert here), but that would leave my GPU idle until Astropulse units start flowing again. Is there any way to distinguish between longer and shorter WUs?

Here's my setup:

11/03/2015 08:51:59 | | CAL: ATI GPU 0: AMD Radeon HD 6520G/6530D/6550D/6620G (SuperSumo) (CAL version 1.4.1848, 512MB, 479MB available, 960 GFLOPS peak)
11/03/2015 08:51:59 | | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 6520G/6530D/6550D/6620G (SuperSumo) (driver version 1642.5 (VM), device version OpenCL 1.2 AMD-APP (1642.5), 512MB, 479MB available, 960 GFLOPS peak)
11/03/2015 08:51:59 | | OpenCL CPU: AMD A8-3850 APU with Radeon(tm) HD Graphics (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1642.5 (sse2), device version OpenCL 1.2 AMD-APP (1642.5))
11/03/2015 08:51:59 | | Processor: 4 AuthenticAMD AMD A8-3850 APU with Radeon(tm) HD Graphics [Family 18 Model 1 Stepping 0]
11/03/2015 08:51:59 | | 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 cx16 popcnt syscall nx lm svm sse4a osvw ibs skinit wdt page1gb rdtscp 3dnowext 3dnow
11/03/2015 08:51:59 | | OS: Microsoft Windows 7: Home Premium x64 Edition, Service Pack 1, (06.01.7601.00)
5) Message boards : Number crunching : Long Seti@home v7.7.03 units (not Astropulse) on AMD GPU - don't finish? (Message 1611800)
Posted 10 Dec 2014 by Profile Graham ThomasProject donor
Thanks for all that helpful advice. I'll take a look at that config file and its help file and see if I can work out what those parameters mean. I'm not sure I want to go for any of the more specialised apps at this point.
6) Message boards : Number crunching : Long Seti@home v7.7.03 units (not Astropulse) on AMD GPU - don't finish? (Message 1611639)
Posted 10 Dec 2014 by Profile Graham ThomasProject donor
Recently I've been getting Seti@Home v7.7.03 workunits that run on my AMD GPU (as opposed to the v7.7.00 units that run on the CPU). They seem to come in two flavours. One has an estimated time of just over an hour, and these units complete with no problems. The other has an estimated time of about three and a half hours, and these units seem to get stuck. Generally, after starting OK, when they get stuck they show as being about 20% done, and although the 'elapsed' column keeps counting the 'remaining' column doesn't change.

The first few times this happened I let the computation proceed for 6-8 hours before aborting. Now I abort as soon as I see them in the queue.

Has anyone else seen this behaviour? Is there anything I can do about it? And are there any negative consequences of aborting units frequently?

Astropulse units complete OK.

Here's my setup:

06/12/2014 11:53:00 | | CAL: ATI GPU 0: AMD Radeon HD 6520G/6530D/6550D/6620G (SuperSumo) (CAL version 1.4.1848, 512MB, 479MB available, 960 GFLOPS peak)
06/12/2014 11:53:00 | | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 6520G/6530D/6550D/6620G (SuperSumo) (driver version 1573.4 (VM), device version OpenCL 1.2 AMD-APP (1573.4), 512MB, 479MB available, 960 GFLOPS peak)
06/12/2014 11:53:00 | | OpenCL CPU: AMD A8-3850 APU with Radeon(tm) HD Graphics (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1573.4 (sse2), device version OpenCL 1.2 AMD-APP (1573.4))
06/12/2014 11:53:00 | | Host name: -----
06/12/2014 11:53:00 | | Processor: 4 AuthenticAMD AMD A8-3850 APU with Radeon(tm) HD Graphics [Family 18 Model 1 Stepping 0]
06/12/2014 11:53:00 | | 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 cx16 popcnt syscall nx lm svm sse4a osvw ibs skinit wdt page1gb rdtscp 3dnowext 3dnow
06/12/2014 11:53:00 | | OS: Microsoft Windows 7: Home Premium x64 Edition, Service Pack 1, (06.01.7601.00)

Graham
7) Message boards : Number crunching : Seti@Home 7.7.03 tasks seem to run on GPU - is this normal? (Message 1598545)
Posted 8 Nov 2014 by Profile Graham ThomasProject donor
Thanks to both of you. I'll (probably) let the scheduler sort itself out.

The progress bar and 'time remaining' indicators don't seem to work well with the v7.7.03 Seti@Home units that are running on the GPU. They don't appear to get done any faster than the others so I'm not sure that there's any advantage in running them in this way.

I'll see how things settle down and then decide whether to intervene.
8) Message boards : Number crunching : Seti@Home 7.7.03 tasks seem to run on GPU - is this normal? (Message 1598516)
Posted 8 Nov 2014 by Profile Graham ThomasProject donor
It used to be the case that my normal Seti@Home tasks ran on my CPU(s) while Astropulse tasks ran on "0.128 CPUs + 1 ATI GPU". (I've got an AMD A8 3850 quad processor with built-in ATI GPU.)

Recently I've been receiving Seti@Home v7.7.03 tasks that claim to be running on 0.128 CPUs + 1 ATI GPU. These tasks all have either (opend_ati_cat132) or (opend_ati5_nocal) following their version number. v7.7.00 tasks are running on CPUs as normal and have nothing in brackets following their version number.

Is this all OK? And if I get more Astropules tasks will these take priority over the Seti@Home tasks on the GPU?
9) Message boards : Number crunching : AP 7 7.04 units download, start, then go to 'waiting to run' (Message 1585717)
Posted 12 Oct 2014 by Profile Graham ThomasProject donor
jschoepf - it sounds like your problem is the same as (or very similar to) the one I had. See which graphics drivers you're using and whether you can update them.

In the meantime, from BOINC/Projects tab/SETI@home, you can go to "Your account", then to SETI@home settings and change them so that your computer does not receive Astropulse v7 units. You may have to untick the box that says something like 'Accept other kinds of workunits if none of those selected are available'. That should stop you getting more and more units until you fix your problem.

Graham
10) Message boards : Number crunching : AP 7 7.04 units download, start, then go to 'waiting to run' (Message 1585715)
Posted 12 Oct 2014 by Profile Graham ThomasProject donor
Claggy - you're right. The SETI@home site is now displaying the new driver information. Thanks again.

Graham
11) Message boards : Number crunching : AP 7 7.04 units download, start, then go to 'waiting to run' (Message 1585375)
Posted 11 Oct 2014 by Profile Graham ThomasProject donor
OK - AMD display driver(s) updated and Astropulse 7 7.04 unit now happily running.

The SETI@home site still shows the old driver on the 'Your Computers' page (I'm not sure if that info will eventually be updated automatically). That's where I took the information from, so if my driver had ever autoupdated it might not have been quite so old. But it was clearly too old for v7 units. If anyone's interested, this is what my BOINC event log now tells me I'm using:

11/10/2014 19:11:51 | | CAL: ATI GPU 0: AMD Radeon HD 6520G/6530D/6550D/6620G (SuperSumo) (CAL version 1.4.1848, 512MB, 479MB available, 960 GFLOPS peak)
11/10/2014 19:11:51 | | OpenCL: AMD/ATI GPU 0: AMD Radeon HD 6520G/6530D/6550D/6620G (SuperSumo) (driver version 1573.4 (VM), device version OpenCL 1.2 AMD-APP (1573.4), 512MB, 479MB available, 960 GFLOPS peak)
11/10/2014 19:11:51 | | OpenCL CPU: AMD A8-3850 APU with Radeon(tm) HD Graphics (OpenCL driver vendor: Advanced Micro Devices, Inc., driver version 1573.4 (sse2), device version OpenCL 1.2 AMD-APP (1573.4))


Thanks everyone.

Graham
12) Message boards : Number crunching : AP 7 7.04 units download, start, then go to 'waiting to run' (Message 1585355)
Posted 11 Oct 2014 by Profile Graham ThomasProject donor
Yes - CPU units are running.

Graham
13) Message boards : Number crunching : AP 7 7.04 units download, start, then go to 'waiting to run' (Message 1585353)
Posted 11 Oct 2014 by Profile Graham ThomasProject donor
Thanks Claggy. I'll see if I can update the driver.

Graham
14) Message boards : Number crunching : AP 7 7.04 units download, start, then go to 'waiting to run' (Message 1585350)
Posted 11 Oct 2014 by Profile Graham ThomasProject donor
Ah - I've just seen there's a message when I hover the mouse over a 'Waiting to run' unit. It says, "Scheduler wait: unsupported OpenCL runtime".

Anything I can do about that? Or should I just abort the units and turn off acceptance of AP v7 until this particular version of OpenCL is supported?

Graham
15) Message boards : Number crunching : AP 7 7.04 units download, start, then go to 'waiting to run' (Message 1585347)
Posted 11 Oct 2014 by Profile Graham ThomasProject donor
Apologies if I should have added this to the 'AP7' thread, but that thread looks like it's dealing with different issues.

In the last day or so I've had a lot of downloads of AP 7 7.04. They start to run, but then (after about a second) they straightaway go into 'Waiting to run'. The next time the project updates, BOINC sees that I have no AP units running on the GPU and downloads some more.

I can tell SETI@home not to accept v7 units, but I'd prefer to find out how to get them to run. What can I do?

I'm running stock Windows BOINC 7.2.42 (x64) on an AMD CPU with built-in Radeon GPU. Specs are as follows:

AuthenticAMD
AMD A8-3850 APU with Radeon(tm) HD Graphics [Family 18 Model 1 Stepping 0]
(4 processors)

AMD AMD Radeon HD 6520G/6530D/6550D/6620G (SuperSumo) (512MB) driver: 1.4.1417 OpenCL: 1.1

I've checked my preferences and have tried restarting BOINC, and then restarting the PC. Still no joy.

Graham
16) Message boards : Number crunching : Multibeam WU appears to end processing but doesn't stop running - general problem? (Message 822048)
Posted 22 Oct 2008 by Profile Graham ThomasProject donor
Thanks. That's worth a try.

Graham
17) Message boards : Number crunching : Multibeam WU appears to end processing but doesn't stop running - general problem? (Message 821899)
Posted 22 Oct 2008 by Profile Graham ThomasProject donor
Every so often (but twice in the past two days, which is why I'm writing) I get a Seti work unit that seems to reach its conclusion (either 100% or 99.997% complete, and no more graphics action) but never stops running. Normally my WUs take either just over an hour or just over four hours, but one of these rogues can go on for over 24 hours.

So I abort, but then of course don't get any credit. Is this a general problem or is it something to do with my setup? (I'm running an Athlon 64 3200 - I know AMD chips take longer to process Astropulse units but I haven't seen any issues relating to multibeam ones.) Should I abort or will the unit eventually complete?

Graham
18) Message boards : Number crunching : BOINC 4 and Win98 (Message 24108)
Posted 9 Sep 2004 by Profile Graham ThomasProject donor
Would it be possible for the BOINC Webmaster(s) to put a notice about the Win98 issues and the availability of 4.08 or newer versions on the home page, so that it stays there until it's deliberately taken down?

The original notice has now scrolled off the first screen, but people running pre-XP Windows are clearly still downloading 4.05 and having their computers crash - see the 'Questions & Problems / Windows' forum. Also, people are still asking for the location of the new alpha versions, which means that kind people in the forums have to keep answering the same question.

Graham
------
Graham Thomas
Innovation Studies
University of East London, UK
19) Message boards : Number crunching : BOINC 4 and Win98 (Message 20878)
Posted 1 Sep 2004 by Profile Graham ThomasProject donor
I've had the Win98 problems with 4.05 too, and don't really have anything to add on the main issues. One thing I've seen that nobody's yet mentioned, though:

After my first 4.05 crash, every time I started BOINC 4.05 (I don't do this any more!) I was getting messages from ZoneAlarm saying 'boinc_gui is trying to act as a server' and 'boinc_gui is trying to access the trusted zone'. I didn't get this from 3.20, and I didn't get it from 4.05 when I installed it (i.e. before the first upload and the first triggering of the slowdown/crash problem).

Allowing or denying this access didn't seem to make any difference to Boinc's behaviour.

I'm hoping this will either turn out to be normal v4 practice or that it will go away when I download v4.06 or higher (I'm going to wait a little while before doing this, as I don't have much time to devote to Boinc right now).

Anyone else seen anything similar?
Graham Thomas
Innovation Studies
University of East London, UK
20) Questions and Answers : Getting started : Can't attach to project - account problem or server problem? (Message 2680)
Posted 29 Jun 2004 by Profile Graham ThomasProject donor
I'm up and running at home, but when I tried just now (Tues 3.30pm in UK) to attach from my work machine I got 'BOINC failed to log in.. Check your account ID' in the alert box, and 'Message from server - server has software problem' in Messages. So, is the problem at my end or the server's? I'm pretty sure I'm pasting in the URL & account ID correctly. I don't think I need to use a proxy from here - I didn't need it for SAH 'classic'. Any suggestions?

Graham Thomas
Innovation Studies
University of East London, UK


Copyright © 2015 University of California