Posts by Raistmer


log in
1) Message boards : Number crunching : GPU performance problems (Message 1705222)
Posted 3 days ago by Profile Raistmer
Thank you all for your help.

Louis Loria II, the thread is very interesting, but I was not able to find anything of worth to solve my particular problem.

Mike, I had change the settings and dedicated 1 CPU to feed GPU, apparently nothing happend, now it's after 92 GPU hours and estimated time to finish is 338 hours. I'm afraid this is not the solution to the problem or I made a mistake...

Check system event log for driver restart event.
2) Message boards : Number crunching : App for Intel GPU on Linux. When? Or never? (Message 1704730)
Posted 5 days ago by Profile Raistmer

And it's not very nice to refer to a thread hidden in some dev only secret lair.

Richard has access.

P.S. And if you have working build bother to share it with the rest, that would be nice.
3) Message boards : Number crunching : App for Intel GPU on Linux. When? Or never? (Message 1704701)
Posted 5 days ago by Profile Raistmer
Juha has got it running at Einstein with beignet.

Einstein message 141681


I think on SETI forum asking for SETI app, not for Einstein's or ABCDEF one.
Look corresponding thread on Lunatics where Urs precisely described current state of affairs.
4) Message boards : Number crunching : to many empty tasks (Message 1704688)
Posted 5 days ago by Profile Raistmer
1) Link only for private access.
2) Where in log 5000 tasks? Log just say that servers off at moment you tried to connect.
5) Message boards : Number crunching : App for Intel GPU on Linux. When? Or never? (Message 1704687)
Posted 5 days ago by Profile Raistmer
Any thoughts?

When OpenCL runtime will work there.
6) Message boards : Number crunching : Monster GPU Cruncher Build (Message 1704433)
Posted 6 days ago by Profile Raistmer
1 Xeon E5-2630v2 have 40 PCIe lanes - I have 2 installed.


And how many each installed GPU takes?...
7) Message boards : Number crunching : New cruncher problems (Message 1704226)
Posted 6 days ago by Profile Raistmer
it can be result of fily system corruption on unexpected previous reboot.
8) Message boards : Number crunching : AMD CPU tips (Message 1704209)
Posted 6 days ago by Profile Raistmer
Been a while since I posted to this thread. To get back to main focus of improving AMD performance versus the status quo of Intel processors, I hope everyone running an AMD FX processor has installed the required Microsoft AMD Scheduler KB's. They are not automatically installed by Windows Update. It is up to the AMD user to install them manually. KB2645594 corrects Windows 7 process scheduler to allow the AMD core module architecture to work properly. If you hadn't installed this patch you wouldn't have been using all your cores. Along with KB2645594 you also need to install KB2646060 which disables core parking which again could make Windows not use all your FX cores. Just wanted to make that fact known since it hadn't come up in the thread before.


An update is available for computers that have an AMD FX, AMD Opteron 4200, AMD Opteron 6200, or AMD Bulldozer series processor installed and that are running Windows 7 or Windows Server 2008 R2


I have some question regarding this topic: can A-10 and C-60 APUs be considered as "AMD Bulldozer series processor" regarding this issue ?
9) Message boards : Number crunching : New cruncher problems (Message 1704201)
Posted 6 days ago by Profile Raistmer
Second reason (besides device overheating) could be device underpowering. PSU could have issues to power CPU working @max power consumption.
10) Message boards : Number crunching : Comparing the performance of various MB apps (Message 1704135)
Posted 7 days ago by Profile Raistmer

BTW: Does anybody know if Cuda will still be officially supported in the future or if MB will switch from Cuda to open CL completly? (openCL app already at beta)


"CUDA officially supported" can be only by nVidia corporation. Only nVidia can either support CUDA in their drivers or drop support. All others can either use CUDA if it supported for particular hardware/OS or use something else. Just to make words little clearer.
11) Message boards : Number crunching : Comparing the performance of various MB apps (Message 1704131)
Posted 7 days ago by Profile Raistmer
The figures on the application details page are averages over recent workunits. Maxwell-based NVidia cards, like the GTX 750 in that host, are notably inefficient at running VHAR 'shorty' tasks. A run of shorties would show up as tasks running longer than estimated, or conversely the card running slower. If you want to compare application speeds accurately, you have to use the same mix of tasks for each application, usually in an off-line bench test.

Another way with less processing overhead (but perhaps requiring more operators work) is to plot time vs AR curve for each of builds. This way real data can be used (online). And more apparent picture of relative performance of builds in comparison can be achieved.
12) Message boards : Number crunching : Windows 10 - Yea or Nay? (Message 1703916)
Posted 7 days ago by Profile Raistmer
Interesting, if compatibility still keeped on modern PCs? Can modern PC with EFI boot into MS-DOS ?...
13) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1703702)
Posted 8 days ago by Profile Raistmer
BTW, speaking about technical consequencies worth to mention top + low-end same vendor GPUs host config.
In such config "re-scheduling" happens on task by task 24/7 basis cause as was said huge number of time BOINC had not device-oriented design. Being non device-oriented it must be fault-tolerant to such configs via broadening acceptable range of deviations from expected values.
We know that low-end GPU devices roughly similar if not slower than modern CPU core.
Hence, there should be no distinction in CPU+GPU with re-scheduling versus low-end GPU + top GPU configs. re-scheduling between CPU core and GPU can't deceive BOINC estimation mechanism more than host with low-end + top GPU (not nessesary top, just low-end GPU and "average" GPU will still fit into my logic).
And hence, either BOINC stable versus re-scheduling or it just can't handle naturally occuring hosts (that, obviously, requires fix).
14) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1703695)
Posted 8 days ago by Profile Raistmer

If you can stand before the angry mob and say 'yes I did it, I had my reasons' - well. But don't be amazed if you get lynched before you managed to explain your reasons.


LoL, I thought westerns became more civilized since those times, that illusion your mass media try to embed in our minds :P

Regarding topic - time estimates as screwed as credit are, does anybody not know this since last app updates?
Anyway host-related effect if any will occur much sooner than project-wide so no reason to vary anyway. Also, you mentioned credit lowering as result not as putative but as fact. Links, please.
In general, moving task from one device to another roughly similar to change properties of device itself.
Few examples: CPU overclocking, GPU underclocking. If underclocking too big indeed, task could be aborted due to not meeting BOINC's expectations for its duration. Well, even taking too much vitamins can kill you as any other mindless action. Всё хорошо в меру (~Measure is a treasure) they say.
15) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1703685)
Posted 8 days ago by Profile Raistmer
So you will still use the rescheduler until the SETI admins post in the News: 'please don't use the rescheduler' ?


I will use it while and when see to appropriate and beneficent for the project.
All objections regarding its usage come from quite few SETI forums members and all based on its possible influence on CreditNew credit awarding scheme that has quite a little connection with things that can be considered as positive for the SETI@home project. It was discussed many times already, I don't see any reason to start new round. Just don't deceive public with from where roots of buzz grow.


I can see that. On a technical level, well you won't find the problems/weaknesses if you don't push the mechanism beyond its limits either. It's tough though IMO. Which is better ? to push a system to the point it collapses, or try to prop it up by good behaviour. Maybe it isn't black and white.

Yep, exactly.
16) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1703679)
Posted 8 days ago by Profile Raistmer
So you will still use the rescheduler until the SETI admins post in the News: 'please don't use the rescheduler' ?


I will use it while and when see to appropriate and beneficent for the project.
All objections regarding its usage come from quite few SETI forums members and all based on its possible influence on CreditNew credit awarding scheme that has quite a little connection with things that can be considered as positive for the SETI@home project. It was discussed many times already, I don't see any reason to start new round. Just don't deceive public with from where roots of buzz grow.
17) Message boards : Number crunching : GPU crunching problem continues.... (Message 1703676)
Posted 8 days ago by Profile Raistmer
So, my HD7870 (BrettPC)was having major issues with tasks, producing a very large amount of Validation inconclusive and invalid's. It was also having other issues and was running at 92-93C under load. It had other issues as well, for example when I plugged in a second monitor to the HDMI port, it crashed windows, and again when I unplugged it.

So I decided to replace it. Installed an MSI HD R9-380, which ran great for 48 hours. Seemed about 25% faster than the HD7870.

Today I get home from work to find a V7 7.07 (opencl_ati5_nocal) WU hung at 6.527% and running for a couple of hours (they usually take anywhere from 5-10 minutes).

I suspended it and tried another, and another, and another........
I rolled the driver back from the latest AMD one to the one that is on the MSI website.

Restarted and seemed to be ok. After a couple of three successes, same thing again.

At the point of hanging (6.5275), I get a windows message saying driver has stopped working and then recovers (screen goes blank for a second.) The card is plugged into the same pciex16 slot as the 7870 was. SO, beginning to think maybe an MB issue? Windows 8.1 Pro 64bit.

Any thoughts, ideas.

Thanks,

Brett


-period_iterations_num 100 -sbs 256
18) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1703672)
Posted 8 days ago by Profile Raistmer
But as Hal said, it may not work for V7 anyway.


Well, again. It works. Personally veryfied. Period.
19) Message boards : Number crunching : Panic Mode On (98) Server Problems? (Message 1703671)
Posted 8 days ago by Profile Raistmer
I myself wrote:
AFAIK, also the SETI admins were 'not pleased' about this tool since CreditNew.

Raistmer wrote:
Link, please.

I was in E-Mail contact with David, Eric, Matt and Jeff.
I asked why they don't write it on the homepage - that not to use a rescheduler tool.

Eric answered: I could make such a statement, but I doubt it would do much good. We have no way of verifying who is using the tool. (...)

So they will not make an officially statement about 'not to use a rescheduler tool'.


So that's better keep clear. All that buzz over re-sheduler comes from forums and forums only.
20) Message boards : Number crunching : Bad News on BOINC funding (Message 1703670)
Posted 8 days ago by Profile Raistmer
SETI@home already got very much. Such big coverage in news over whole world! I think it was never happened before in such scale.


Next 20

Copyright © 2015 University of California