Advanced search

Message boards : Graphics cards (GPUs) : CPU request is now

Author Message
Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 3267 - Posted: 25 Oct 2008 | 11:06:01 UTC
Last modified: 29 Oct 2008 | 8:45:00 UTC

GPUGRID is now requesting less than one processor such that people can run a further cpu job.
This will stay if there are no problems. People should not use ncpus anymore and update to clients 6.3.19.

GDF

Profile ayQue
Send message
Joined: 6 Sep 08
Posts: 18
Credit: 806,771
RAC: 0
Level
Gly
Scientific publications
watwatwatwat
Message 3268 - Posted: 25 Oct 2008 | 11:15:01 UTC

Nice! I'll try this later @ home and report... :)

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3269 - Posted: 25 Oct 2008 | 12:13:40 UTC - in response to Message 3267.

GPUGRID is now requesting less than one processor such that people can run a further cpu job.
This will stay if there are no problems. People should not use ncpus anymore and update to clients 6.3.17 or 6.3.14.

GDF

I would say use 6.3.17. So far no problems. I did extensive testing with 6.3.15 with no problems as far as job scheduling, testing both CPU=1 and CPU<1 jobs, which always ran and ran the proper number of other CPU only jobs. 6.3.17 should be the same as far as that goes.

6.3.14 will work some, but eventually it will leave no GPU tasks running, or only a GPU running. It does not behave well.

Profile UL1
Send message
Joined: 16 Sep 07
Posts: 56
Credit: 35,013,195
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwat
Message 3270 - Posted: 25 Oct 2008 | 12:39:56 UTC

Any idea when 6.3.17 for Linux64 will be out ? 6.3.14 won't work for me...

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3272 - Posted: 25 Oct 2008 | 12:48:14 UTC - in response to Message 3270.

Any idea when 6.3.17 for Linux64 will be out ? 6.3.14 won't work for me...

No I don't, but usually once the windows version is released to alpha, which is was on Friday, the linux version follows in a few days. Being that it is now the week-end I would guess not at least until Monday.

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3287 - Posted: 25 Oct 2008 | 17:39:15 UTC - in response to Message 3267.
Last modified: 25 Oct 2008 | 17:40:36 UTC

Hehe, yea it's now <1...
BOINC shows 0.90 CPUs, 1 CUDA.

BOINC also shows 4 CPU tasks running plus 1 PS3GRID, but the taskmanager shows:

3 Einstein tasks with each 25% CPU
1 Cosmo task with 25%
and all few seconds I can see the acemd task kicking in with 1 or 2% CPU.

Is this how it should work?
I would think with 0.9 CPU the PS3GRID task should get more of the CPU?

[edit]Using 6.3.17, of course without ncpus+1[/edit]
____________

pixelicious.at - my little photoblog

fractal
Send message
Joined: 16 Aug 08
Posts: 87
Credit: 1,248,879,715
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3290 - Posted: 25 Oct 2008 | 18:33:45 UTC

Do we need to wait for new work for this to take affect?

I am seeing "Running (1.00 CPUs, 1 CUDA)" on 6.3.17

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 3292 - Posted: 25 Oct 2008 | 19:15:32 UTC - in response to Message 3287.

0.9 is only indicative, as long is less than 1.

gdf

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 3293 - Posted: 25 Oct 2008 | 19:16:08 UTC - in response to Message 3292.

Please check ms/step to see if they change.

gdf

Profile [SETI.USA]Tank_Master
Avatar
Send message
Joined: 8 Jul 07
Posts: 85
Credit: 67,463,387
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 3296 - Posted: 25 Oct 2008 | 19:22:00 UTC - in response to Message 3287.

Hehe, yea it's now <1...
BOINC shows 0.90 CPUs, 1 CUDA.

BOINC also shows 4 CPU tasks running plus 1 PS3GRID, but the taskmanager shows:

3 Einstein tasks with each 25% CPU
1 Cosmo task with 25%
and all few seconds I can see the acemd task kicking in with 1 or 2% CPU.

Is this how it should work?
I would think with 0.9 CPU the PS3GRID task should get more of the CPU?

[edit]Using 6.3.17, of course without ncpus+1[/edit]


my task manager shows 3 ramsey WUs at 25%, 1 at 17-18% and the GPU app at 7-8%

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3301 - Posted: 25 Oct 2008 | 19:49:29 UTC - in response to Message 3287.
Last modified: 25 Oct 2008 | 19:55:15 UTC

Hehe, yea it's now <1...
BOINC shows 0.90 CPUs, 1 CUDA.

BOINC also shows 4 CPU tasks running plus 1 PS3GRID, but the taskmanager shows:

3 Einstein tasks with each 25% CPU
1 Cosmo task with 25%
and all few seconds I can see the acemd task kicking in with 1 or 2% CPU.

Is this how it should work?
I would think with 0.9 CPU the PS3GRID task should get more of the CPU?

[edit]Using 6.3.17, of course without ncpus+1[/edit]

Yes, that is correct behavior.

The 0.9 can be any number less that 1.

If you check all CPU only jobs will run at low system priority.
CUDA/CPU=1 will run at low system priority
CUDA/CPU<1 will run at normal system priority.
This is for windows, i do not know how linux call it, but CPU<1 should run at a higher priority than the others.

In the future when there are more projects and more GPU's the CPU<1 percentage will matter, but for now with only one project it does not, as long as it is less than 1. I think the goal will be to use up to physical CPUs + 0.99

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3302 - Posted: 25 Oct 2008 | 19:51:23 UTC - in response to Message 3290.

Do we need to wait for new work for this to take affect?

I am seeing "Running (1.00 CPUs, 1 CUDA)" on 6.3.17

Yes, any exsiting work you have will run at what ever it was sent at. You will notice the change was new work starts to download, although there may still be some work in the server queue at the old setting.

Profile Wassertropfen
Avatar
Send message
Joined: 14 Aug 08
Posts: 15
Credit: 13,774,919
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3303 - Posted: 25 Oct 2008 | 20:01:23 UTC

The 6.3.17 is working fine.

One Problem with vista 64. Without admin rights only three task + ps3grid are running on a quad. With admin rigts there are 4 task + ps3grid running.

any help?
____________
Constant dripping wears away the stone. :)

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3304 - Posted: 25 Oct 2008 | 20:06:00 UTC - in response to Message 3301.


Yes, that is correct behavior.

The 0.9 can be any number less that 1.

If you check all CPU only jobs will run at low system priority.
CUDA/CPU=1 will run at low system priority
CUDA/CPU<1 will run at normal system priority.
This is for windows, i do not know how linux call it, but CPU<1 should run at a higher priority than the others.

In the future when there are more projects and more GPU's the CPU<1 percentage will matter, but for now with only one project it does not, as long as it is less than 1. I think the goal will be to use up to physical CPUs + 0.99


Ok...
I'm only looking at the Vista host now, because it is more critical.
On the Linux hosts I still use ncpus+1 until the new setting hits them too...


I just checked the priority and the PS3GRID task was running at the lowest priority. After restarting BOINC the acemd task is now running at normal priority, and still shows the same behaviour. CPU only tasks get ~25% (one full core) and ps3grid only gets one or two percent CPU all few seconds...
Let`s see if the ms/step changes after the first few results...
____________

pixelicious.at - my little photoblog

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3305 - Posted: 25 Oct 2008 | 20:06:37 UTC - in response to Message 3303.

The 6.3.17 is working fine.

One Problem with vista 64. Without admin rights only three task + ps3grid are running on a quad. With admin rigts there are 4 task + ps3grid running.

any help?

Get rid of vista.

Sorry its been a long day...

No, I don't know about that one. I'll inquire on the alpha testers list.

ps, for those with the startup wizard attach problem, i've inquired on that too.

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3306 - Posted: 25 Oct 2008 | 20:08:27 UTC - in response to Message 3303.
Last modified: 25 Oct 2008 | 20:10:28 UTC

Admin rights shouldn't be necessary!

I have 4 CPU tasks plus one GPU running on my Vista64 box. Just wait a little bit until the next task switch, the changes don't happen immediately. Or maybe restart BOINC...
____________

pixelicious.at - my little photoblog

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3308 - Posted: 25 Oct 2008 | 20:17:04 UTC - in response to Message 3306.

Admin rights shouldn't be necessary!

I have 4 CPU tasks plus one GPU running on my Vista64 box. Just wait a little bit until the next task switch, the changes don't happen immediately. Or maybe restart BOINC...

That is what I would think too, but it should kick in another task right away. I have not seen since 6.3.15 any improper number of tasks running on three windows XP clients, always immediately another one will start, or download more work if necessary and start. I've tried many things to break the scheduler and have not been able to. I've tired many combination of projects, cache size, cpus, suspended tasks, suspended project, and so on. Most always it will change immediately or within 2 seconds at most.

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3309 - Posted: 25 Oct 2008 | 20:21:00 UTC
Last modified: 25 Oct 2008 | 20:45:02 UTC

Now for Linux I have a Q9550 running 4 cpu tasks and the gpu is always waiting to run 0.90 cpu,1 CUDA ????It starts up correctly but in 30 seconds or so it switches the CUDA app to waiting to run. Tried BOINC 6.3.10 & 6.3.14 same effect, Also the system monitor is showing only about 80% of resources are being used.My other Q9550 box is running correctly 4 cpu's and 1 0.90cpu,1CUDA for 5 total threads and 100% utilization...

What happenned? What can I do?

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3311 - Posted: 25 Oct 2008 | 20:23:10 UTC - in response to Message 3308.

Hmm... Now I also have only 3 CPU tasks and one GPU task running...

I'll set up cc_config with a few debug flags tomorrow morning to set what's happening.
____________

pixelicious.at - my little photoblog

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3316 - Posted: 25 Oct 2008 | 21:13:05 UTC - in response to Message 3309.

Now for Linux I have a Q9550 running 4 cpu tasks and the gpu is always waiting to run 0.90 cpu,1 CUDA ????It starts up correctly but in 30 seconds or so it switches the CUDA app to waiting to run. Tried BOINC 6.3.10 & 6.3.14 same effect, Also the system monitor is showing only about 80% of resources are being used.My other Q9550 box is running correctly 4 cpu's and 1 0.90cpu,1CUDA for 5 total threads and 100% utilization...

What happenned? What can I do?

Nothing. Client flaws.

In my testing (all on windows), the CPU<1 only works correctly in 6.3.15, 6.3.16 and 6.3.17.

6.3.14 and 6.3.10 had flaws which prevented the correct behavior. Sometimes it might work, but more often not, leaving less than max cpus or no gpu task running.

@stefan
Which version client are you running ?

@GDF
Maybe you should wait until an official 6.3.17 linux is made, as the last one I see is 6.3.14 on the dl page. Maybe you should switch back to CPU=1 until one is released, unless you have one all linux users can download.

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3318 - Posted: 25 Oct 2008 | 21:23:48 UTC - in response to Message 3316.

@ Keith - 6.3.17

Now I have four CPU and one GPU task running again. But it's to late for me to set up the debug stuff and watch BOINC... ;) Will do that tomorrow...
____________

pixelicious.at - my little photoblog

Profile koschi
Avatar
Send message
Joined: 14 Aug 08
Posts: 124
Credit: 466,579,198
RAC: 297,582
Level
Gln
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 3321 - Posted: 25 Oct 2008 | 21:49:51 UTC - in response to Message 3316.

@GDF
Maybe you should wait until an official 6.3.17 linux is made, as the last one I see is 6.3.14 on the dl page. Maybe you should switch back to CPU=1 until one is released, unless you have one all linux users can download.


++

Since 0.9 CPU setting is active, my GPU stops crunching after restarting the client within few minutes.
We should wait for 6.3.17 to be available for Linux also, to enable the 0.9 CPU share/weight/whatever...

STE\/E
Send message
Joined: 18 Sep 08
Posts: 368
Credit: 315,034,798
RAC: 555,655
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 3322 - Posted: 25 Oct 2008 | 21:50:48 UTC

So far except for a very brief period all I've ever seen on my Box's is 4 Wu's running. On 1 there were 5 running for about 5 minutes & then that was it, 4 on that Box too ever since, running 6.3.17 on all 5 of my GPU CUDA capable Box's.

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3324 - Posted: 25 Oct 2008 | 22:02:13 UTC - in response to Message 3318.

@ Keith - 6.3.17

Now I have four CPU and one GPU task running again. But it's to late for me to set up the debug stuff and watch BOINC... ;) Will do that tomorrow...

Is this on your windows vista host ?

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3325 - Posted: 25 Oct 2008 | 22:06:47 UTC - in response to Message 3322.

So far except for a very brief period all I've ever seen on my Box's is 4 Wu's running. On 1 there were 5 running for about 5 minutes & then that was it, 4 on that Box too ever since, running 6.3.17 on all 5 of my GPU CUDA capable Box's.

Can you specify the mix of 4 and 5 ?

What type of boxes ? Since your computers are hidden I can't look it up.

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3327 - Posted: 25 Oct 2008 | 22:12:30 UTC - in response to Message 3324.

Yup Keith, that's on the Vista Ultimate 64 bit box, with BOINC 6.3.17.
____________

pixelicious.at - my little photoblog

frankhagen
Send message
Joined: 18 Sep 08
Posts: 65
Credit: 3,037,414
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 3328 - Posted: 25 Oct 2008 | 22:45:41 UTC - in response to Message 3327.

hmm - i'm running fiesta x64u too (6.3.17) - 4 ABC-WUs and 1 Cuda - the only thing is that 1 of the ABC-tasks is suspending from time to time.

Profile paul
Avatar
Send message
Joined: 16 Oct 08
Posts: 7
Credit: 3,999,754
RAC: 0
Level
Ala
Scientific publications
watwatwatwat
Message 3329 - Posted: 25 Oct 2008 | 23:00:15 UTC - in response to Message 3267.

running XP32 SP3, was running fine, 4 3x+1 wu and 1 GPU with ncpu+1. I removed the ncpu+1, installed BOINC 5.3.17, restarted, let it go through a completed gpu wu, it's now running 3 3x+1 and 1 GPU grid wu
____________
Team Starfire World BOINC
IRC- irc//irc.teamstarfire.net:6667/team_starfire

fractal
Send message
Joined: 16 Aug 08
Posts: 87
Credit: 1,248,879,715
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3330 - Posted: 25 Oct 2008 | 23:14:01 UTC - in response to Message 3302.

Do we need to wait for new work for this to take affect?

I am seeing "Running (1.00 CPUs, 1 CUDA)" on 6.3.17

Yes, any exsiting work you have will run at what ever it was sent at. You will notice the change was new work starts to download, although there may still be some work in the server queue at the old setting.
It is still on the same wu, but I have updated the project. I now see



With 4 cpu tasks running, 1 gpu task running (high priority). Unfortunately, taskman shows the GPU task running at the same priority as the CPU task, so is going to fight the cpu.

JKuehl2
Send message
Joined: 18 Jul 08
Posts: 33
Credit: 3,233,174
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 3331 - Posted: 25 Oct 2008 | 23:27:47 UTC
Last modified: 25 Oct 2008 | 23:29:37 UTC

confirmed in my team, all three machines using 6.3.17:

working on two xp 32 bit machines as expected

NOT working on my vista 64 bit machine (instead 16-18% CPU, task gets only 1-2%, CPU Tasks get 24-25% each, slowdown as expected). Therefore im still using ncpus=4 till this is resolved.

fractal
Send message
Joined: 16 Aug 08
Posts: 87
Credit: 1,248,879,715
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3341 - Posted: 26 Oct 2008 | 3:26:30 UTC - in response to Message 3330.

With 4 cpu tasks running, 1 gpu task running (high priority). Unfortunately, taskman shows the GPU task running at the same priority as the CPU task, so is going to fight the cpu.

Well, the WU in question finished and a new one started. It is running the same (0.9 CPU, 1GPU) but the thread is running at NORMAL priority according to taskman. This is what was advertised and what I am seeing, so it looks good.

Profile K1atOdessa
Send message
Joined: 25 Feb 08
Posts: 249
Credit: 370,320,941
RAC: 0
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3342 - Posted: 26 Oct 2008 | 3:45:15 UTC - in response to Message 3269.
Last modified: 26 Oct 2008 | 3:51:18 UTC

GPUGRID is now requesting less than one processor such that people can run a further cpu job.
This will stay if there are no problems. People should not use ncpus anymore and update to clients 6.3.17 or 6.3.14.

GDF

I would say use 6.3.17. So far no problems. I did extensive testing with 6.3.15 with no problems as far as job scheduling, testing both CPU=1 and CPU<1 jobs, which always ran and ran the proper number of other CPU only jobs. 6.3.17 should be the same as far as that goes.

6.3.14 will work some, but eventually it will leave no GPU tasks running, or only a GPU running. It does not behave well.


I had been running 6.3.14, but just moved to 6.3.17. My system is now crunching 2 WUs with .9 CPU 1 CUDA, and after the initial install of 6.3.17 was crunching 3 other CPU tasks. Now, after 1 hour or so, it is back to running on 2 CPU + 2 GPU tasks -- pretty much the same results with 6.3.14 before the .9 CPU 1 CUDA was turned on. I generally see the 2 CPU tasks running 25% each on my Q6600, with the 2 GPU tasks taking about 18% each -- leaving about 12% idle.

UPDATE: Once I closed BOINC manager and the boinc.exe process, reopening the BOINC manager once again restarted 3 CPU tasks + 2 GPU tasks. Within seconds, the extra CPU task went back to "Waiting to run" which is what happened before. Apparently, this is repeatable.

Thamir Ghaslan
Send message
Joined: 26 Aug 08
Posts: 55
Credit: 1,475,857
RAC: 0
Level
Ala
Scientific publications
watwatwat
Message 3346 - Posted: 26 Oct 2008 | 8:12:04 UTC - in response to Message 3342.

I have Boinc .14 that was running 5 tasks set by the ncpu flag.

After noticing the .9 CPU 1 cuda switch the project now runs 5 cpu tasks and 1 GPU.

I've deleted the cc_config file and now it runs 4 tasks plus 1 GPU.

The GPU consumes 2% to 5% of the CPU.

Already noticing the 30% performance increase, my daily 9500 credits jumped to the 13000s.

Bottom line, good work, although I'm hesitant to switch to boinc .17

Profile Wassertropfen
Avatar
Send message
Joined: 14 Aug 08
Posts: 15
Credit: 13,774,919
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3350 - Posted: 26 Oct 2008 | 9:28:43 UTC - in response to Message 3303.

Without admin rights only three task + ps3grid are running on a quad. With admin rigts there are 4 task + ps3grid running.

That is not correct! It was my mistake. I haven't enough time. All is working fine with normal user rights.

But! The most time only 3 Task and 1 cuda running in stet of 4 task and 1 cuda.
____________
Constant dripping wears away the stone. :)

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3354 - Posted: 26 Oct 2008 | 10:02:17 UTC

Klat,

I think I've seen the same behaviour as you did when I switched to 6.3.17 yesterday. BOINC manager showed 3 CPU + 1 GPU task running on my quad. However, there's a minor screen refresh issue where the manager shows a task as "ready to run" whereas in reality it's already started. I checked with task manager: 4 CPU and 1 GPU were running, although the 4th CPU task almost didn't get any CPU time, overall utilization was 93%.

After this night everything is running as expected, though. Didn't change anything. Now it's 4 CPU + 1 GPU, cpu utilization 100% and the GPU task taking between 9 and 22% (one core 25%). GPU temperature seems to be the same so I hope it's still running full throttle.

MrS
____________
Scanning for our furry friends since Jan 2002

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3355 - Posted: 26 Oct 2008 | 10:12:04 UTC - in response to Message 3346.
Last modified: 26 Oct 2008 | 10:12:29 UTC

Already noticing the 30% performance increase, my daily 9500 credits jumped to the 13000s.


You mean your GPU-Grid credits jumped from ~9.500 per day to 13.000?

Well, that's not normally related to the "CPU request is now < 1" switch. You should have gotten the higher amount of credtis since the 6.48 client was released, but at the expense of sacrificing CPU time. If you are seeing the improved GPU times only now it means your GPUs were CPU-limited before and could not reach their potential due to the use of ncpus+1 (or 2 in your case). What helps you is probably the "normal" priority setting, which was introduced now.

Bottom line, good work, although I'm hesitant to switch to boinc .17


Why? (the answer may very well belong into the thread for 6.3.17)

MrS
____________
Scanning for our furry friends since Jan 2002

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3357 - Posted: 26 Oct 2008 | 10:42:53 UTC - in response to Message 3354.

Klat,

I think I've seen the same behaviour as you did when I switched to 6.3.17 yesterday. BOINC manager showed 3 CPU + 1 GPU task running on my quad. However, there's a minor screen refresh issue where the manager shows a task as "ready to run" whereas in reality it's already started. I checked with task manager: 4 CPU and 1 GPU were running, although the 4th CPU task almost didn't get any CPU time, overall utilization was 93%.

After this night everything is running as expected, though. Didn't change anything. Now it's 4 CPU + 1 GPU, cpu utilization 100% and the GPU task taking between 9 and 22% (one core 25%). GPU temperature seems to be the same so I hope it's still running full throttle.

MrS


Well, actually there's really something wrong with the scheduler. I noticed a few times that BOINC sometimes runs 3 CPU plus 1 GPU task, and sometimes 4 CPU plus 1 GPU task. I already reported this to BOINC alpha, but haven't got an answer until now...
____________

pixelicious.at - my little photoblog

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3359 - Posted: 26 Oct 2008 | 10:54:22 UTC

The problem could be related to what I posted in the 6.3.17 thread (resuming issue).

MrS
____________
Scanning for our furry friends since Jan 2002

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3360 - Posted: 26 Oct 2008 | 11:00:48 UTC - in response to Message 3359.

From the logs it looked like it has to do with the preemption of tasks, but I'm not a dev, I only report what I see... ;)
____________

pixelicious.at - my little photoblog

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3363 - Posted: 26 Oct 2008 | 13:45:51 UTC
Last modified: 26 Oct 2008 | 14:07:27 UTC

Would like to report back that under Linux and 6.3.14 the gpu task did eventually start and run full time,but it was a different task as I aborted a few that would only run 30 secs every 30 min...since it is on a 9800GT it is only 75% complete and will report back what happens as it completes to start the next one.

Edit-Bad news...my other host running CUDA on Linux is exhibiting the same behaviour as in waiting to run.....4cpu's running gpu waiting :(...I hope they release a new client soon for Linux...

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3365 - Posted: 26 Oct 2008 | 14:21:46 UTC

Is your ressource share high enough, say >30% than the combined share for all cpu projects?

MrS
____________
Scanning for our furry friends since Jan 2002

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3367 - Posted: 26 Oct 2008 | 14:34:51 UTC - in response to Message 3365.

Yes ETA it is 5-10 times higher than any individual project and over 30% higher than the total ....but I will crank it up even more....good idea

Profile Nightlord
Avatar
Send message
Joined: 22 Jul 08
Posts: 61
Credit: 5,461,041
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3368 - Posted: 26 Oct 2008 | 14:35:48 UTC

For info: I tried experimenting by setting cpu_sched=0 in the cc_config file.

It appears to have stopped boinc pre-empting the GPU task for now and, as yet, I see no other unwanted behaviour.

These two Linux boxes are running 1 CPU project (2 tasks) plus GPU grid, both projects at 50% resource share They run with 6.3.10, and ncpus=2.

____________

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3369 - Posted: 26 Oct 2008 | 14:40:56 UTC

Nightlord could you layout the cc_config file the way you have it setup so that I may copy and paste it please?

Profile Nightlord
Avatar
Send message
Joined: 22 Jul 08
Posts: 61
Credit: 5,461,041
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3371 - Posted: 26 Oct 2008 | 14:46:17 UTC - in response to Message 3369.
Last modified: 26 Oct 2008 | 14:50:31 UTC

This is my cc_config.xml file:


<cc_config>
<log_flags>
<task>1</task>
<cpu_sched>0</cpu_sched>
</log_flags>
<options>
<ncpus>2</ncpus>
<report_results_immediately>1</report_results_immediately>
</options>
</cc_config>


I think there are dangers in adjusting the cpu scheduler setting, but it seems to be ok for one CPU project plus GPU....be careful if you run more than one CPU project.

YMMV

edit: it last lost some formatting by posting to the forum, so you may need to adjust the tabbed spacing in your file.
____________

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3372 - Posted: 26 Oct 2008 | 14:59:21 UTC - in response to Message 3371.

Ok thanks its appreciated...I will adjust it as I am running this on a quad and will watch for strange behaviour...

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3373 - Posted: 26 Oct 2008 | 15:03:13 UTC - in response to Message 3371.

I don't see why a log flag should change the behaviour of the scheduler, but who knows... ;)
____________

pixelicious.at - my little photoblog

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3374 - Posted: 26 Oct 2008 | 15:26:47 UTC

Nope ....good try Nightlord but did not work....on bootup runs 20-30 secs and still goes to waiting.

Checked std and ltd in client state and its massively to the plus side and exponentially greater than any other project.

Profile Nightlord
Avatar
Send message
Joined: 22 Jul 08
Posts: 61
Credit: 5,461,041
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3375 - Posted: 26 Oct 2008 | 15:36:38 UTC
Last modified: 26 Oct 2008 | 16:13:07 UTC

Very odd indeed and I agree with Stefan - why should setting the log flag work? But it has for me......on 3 boxes now!

Maybe it is not the flag, but something else I did in the process. So here is a description of exactly what I did in case it's different:

Leave client and science apps runing.
Edit cc_config file and save overwriting existing file (after taking a backup!).
Suspend each project from the Projects tab.
Disconnect from the active client by using Advanced>Shutdown Connected Client.
Close BoincManager.
Verify Boinc is not running using System Monitor.
Start Boinc.
Start BoincManager.
Resume projects from Projects tab.



Edit: maybe it is the suspend/shutdown/start/resume process?

I just ran the process described above on a Vista machine. This box has no cc_config file and I did not add one. After suspending, shutting down,re-starting and resuming as above, it appears to be working normally: 20 minutes elapse with no pre-empting or switching. Two CPU tasks, plus a GPU unit running.
____________

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3376 - Posted: 26 Oct 2008 | 16:13:45 UTC - in response to Message 3375.

Very odd indeed and I agree with Stefan - why should setting the log flag work? But it has for me......on 3 boxes now!

Maybe it is not the flag, but something else I did in the process. So here is a description of exactly what I did in case it's different:

Leave client and science apps runing.
Edit cc_config file and save overwriting existing file (after taking a backup!).
Suspend each project from the Projects tab.
Disconnect from the active client by using Advanced>Shutdown Connected Client.
Close BoincManager.
Verify Boinc is not running using System Monitor.
Start Boinc.
Start BoincManager.
Resume projects from Projects tab.

The log flag is off by default. You setting it to zero does nothing. If turned on all it does is log messages. It has nothing to do with your edit. Why, because I already have that set off and still am having problems.

By shutting down you have reset cleint and cleared the bug, It will begin to misbehave again when whatever is triggering the problem happens.

You could simply use in boinc manager, advanced > read config file, without shutting down.

Profile Krunchin-Keith [USA]
Avatar
Send message
Joined: 17 May 07
Posts: 512
Credit: 111,288,061
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3377 - Posted: 26 Oct 2008 | 16:18:06 UTC - in response to Message 3367.

Yes ETA it is 5-10 times higher than any individual project and over 30% higher than the total ....but I will crank it up even more....good idea

I find this odd, when was testing before and had a problem running GPU work, I found lowering the resource share to work.

I had something like
A=64
B= 8
GPU=768

When I lowered GPU to 256 it began running again. I've kept it there since and all testing went will, with 6.3.15, with it set lower.

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3378 - Posted: 26 Oct 2008 | 17:02:56 UTC

It has now resumed the task after increasing resource(may or may not have anything to do with it)and should run to completion if my other quad is any indicator....it is at 95% complete now and I suspect it will not start another task immediately but will go to waiting again for a few hours ...will let you know

JKuehl2
Send message
Joined: 18 Jul 08
Posts: 33
Credit: 3,233,174
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 3382 - Posted: 26 Oct 2008 | 17:49:44 UTC

PROBLEM solved:

now working as expected with vista 64 bit. Why? Don´t know - maybe the WUs themselves got some internal logic?

No slowdown anymore, GPU Process stays at 16% average CPU time and the remaining 8-9 are consumed by other projects.

VERY nice job!

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3383 - Posted: 26 Oct 2008 | 18:14:13 UTC

The task completed.....5 minutes later it started another for 30 seconds and is now waiting to run again.No joy in Mudville :(

frankhagen
Send message
Joined: 18 Sep 08
Posts: 65
Credit: 3,037,414
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwat
Message 3386 - Posted: 26 Oct 2008 | 19:46:30 UTC - in response to Message 3383.

now it had me 2 - 1 GPUgrid WU running, 3 on the bench and boincman deciding not to download work for other projects. solved this by cleaning up the bench and increasing work-buffer..

this whole logic is totally fouled up. unless boinc learns to handle resources for different projects seperately, it will never work. :(

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3390 - Posted: 26 Oct 2008 | 21:31:22 UTC

After playing around with my Linux hosts today what I have found for what it is worth is the more task switching you can get the client to do the more liklihood that it will kickstart the gpu task to run to completion....any reboots or client shutdowns will usually put it back to waiting otherwise if it runs more than 30 secs it goes to completion.

I have set my local preferences as low as 5 min to switch and made sure I had multiple projects with work and the gpu downtime is then usually less than 1/2 hour....glad its sunday and not mid-week.

Profile Nightlord
Avatar
Send message
Joined: 22 Jul 08
Posts: 61
Credit: 5,461,041
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3397 - Posted: 26 Oct 2008 | 23:24:35 UTC - in response to Message 3376.
Last modified: 26 Oct 2008 | 23:25:36 UTC

Doh!....Yes of course! Don't think I wasn't thinking straight earlier today....

Anyway, perhaps an important thing which might help trying to locate the source of this problem: it is happening here on 6.3.10 on Linux and Vista32 (I have no rigs higher than this), so tracking it down between 6.3.15 and .17 as suggetsed in the 6.3.17 thread might not be so easy.

I was out of circulation for most of last week: What else changed in between: the app I think?
____________

sigma-7
Send message
Joined: 27 Aug 08
Posts: 3
Credit: 1,608,041
RAC: 0
Level
Ala
Scientific publications
watwatwatwat
Message 3401 - Posted: 27 Oct 2008 | 7:26:44 UTC

Please put back <1 to 1 cpu

with other tasks running [4 seti +1 gpu] the gpu gets starved of cpu time on vista x64

went from 6 hours a wu to over 10 hours with cpu<1 on nvidia 280gtx

it is worth sacrificing 1 cpu for better speed

Profile [BOINC@Poland]AiDec
Send message
Joined: 2 Sep 08
Posts: 53
Credit: 9,213,937
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwat
Message 3405 - Posted: 27 Oct 2008 | 8:30:48 UTC - in response to Message 3401.
Last modified: 27 Oct 2008 | 8:32:14 UTC

Please put back <1 to 1 cpu

with other tasks running [4 seti +1 gpu] the gpu gets starved of cpu time on vista x64

went from 6 hours a wu to over 10 hours with cpu<1 on nvidia 280gtx

it is worth sacrificing 1 cpu for better speed



I agree. But maybe not all of users would like to have the same. Mby best solution will be to give a choice to the ppl? I am thinking about one more functionality in BOINC Manager (to chose <1 or 1). It`s just an idea :).
____________

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 3407 - Posted: 27 Oct 2008 | 8:38:08 UTC - in response to Message 3401.

We have tested this and it should not happen.

gdf

Profile Kokomiko
Avatar
Send message
Joined: 18 Jul 08
Posts: 190
Credit: 24,093,690
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 3409 - Posted: 27 Oct 2008 | 10:23:37 UTC

Same problem here with my GTX280. The real running time is changing on Vista 64bit from 24000 sec to 34800 sec, now my GTX260 is much faster on XP 64 bit. Normally it should be reverse.

GTX280 http://www.ps3grid.net/results.php?hostid=7785 on Vista 64 bit
GTX260² http://www.ps3grid.net/results.php?hostid=13337 on XP 64 bit


____________

Profile [BOINC@Poland]AiDec
Send message
Joined: 2 Sep 08
Posts: 53
Credit: 9,213,937
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwat
Message 3411 - Posted: 27 Oct 2008 | 12:41:13 UTC - in response to Message 3407.

We have tested this and it should not happen.

gdf


It is unfortunatelly happening :(. Too many times my 280 is waiting to run :(.
____________

Sherman H.
Send message
Joined: 28 Sep 08
Posts: 27
Credit: 5,659,777,872
RAC: 4,000,885
Level
Tyr
Scientific publications
watwatwatwat
Message 3412 - Posted: 27 Oct 2008 | 13:08:37 UTC - in response to Message 3407.

We have tested this and it should not happen.

gdf


This also happens on my XP machines, which, for me, can be solved by running the CUDA task at realtime priority. The following is probably close to a worst-case scenario: without modifying Windows priority: 91,256ms/step; realtime priority: 69.999ms/step (I expect it to be around 67.xxx had this task been run at realtime priority throughout).

Sherman H.
Send message
Joined: 28 Sep 08
Posts: 27
Credit: 5,659,777,872
RAC: 4,000,885
Level
Tyr
Scientific publications
watwatwatwat
Message 3413 - Posted: 27 Oct 2008 | 13:10:29 UTC - in response to Message 3401.

Please put back <1 to 1 cpu


Just want to say I myself prefer <1 CPU, since I don't have to muck about with ncpus now. The time sharing for non CUDA project works better without ncpus for me.

Profile Stefan Ledwina
Avatar
Send message
Joined: 16 Jul 07
Posts: 464
Credit: 135,911,881
RAC: 892
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 3417 - Posted: 27 Oct 2008 | 17:53:57 UTC - in response to Message 3407.
Last modified: 27 Oct 2008 | 17:54:17 UTC

We have tested this and it should not happen.

gdf


Sorry GDF, but it does happen...
I also had one of my computers today only crunching 4 CPU tasks but no GPU task.
And there wan nothing in deadline trouble. The 6.3.17 scheduler is still buggy... :(
____________

pixelicious.at - my little photoblog

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 3421 - Posted: 27 Oct 2008 | 22:20:09 UTC - in response to Message 3417.

huumm.
Let's see what they (BOINC) can do in a couple of days. This thing has been under tests for months and was working fine before, if we go back now it is likely to stay on hold for a long time.

gdf

Profile koschi
Avatar
Send message
Joined: 14 Aug 08
Posts: 124
Credit: 466,579,198
RAC: 297,582
Level
Gln
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 3426 - Posted: 27 Oct 2008 | 22:30:46 UTC - in response to Message 3417.
Last modified: 27 Oct 2008 | 22:32:08 UTC

Same for the scheduler in 6.3.18, I described my problem in that thread.
6.3.14 + CPU setting of 1 and cpus +1 was the best solution for me so far, under Linux...
Since we are on 0.9 the scheduler is behaving strange...

edit:

ok, haven't seen your post, so I'll wait ;-)

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3428 - Posted: 27 Oct 2008 | 22:50:42 UTC - in response to Message 3421.

This thing has been under tests for months and was working fine before


The scheduling errors we are currently seeing seem highly random (at least under windows, don't know about this linux problem). For many people it works just as expected, whereas on other machines the scheduler just seems to be crazy.

How would you test for such an error? The old problem of software reliability.. how could you be sure you tested all possible cases, if you can not proove your algorithm mathematically?

The BOINC scheduler seems to be quite complex beast already. Otherwise such *simple* things like "first priority: keep co-processors busy" just wouldn't go wrong, would they?

.. just some random thoughts before going to bed ;)

MrS
____________
Scanning for our furry friends since Jan 2002

Profile Venturini Dario[VENETO]
Send message
Joined: 26 Jul 08
Posts: 44
Credit: 4,832,360
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwatwatwatwat
Message 3445 - Posted: 28 Oct 2008 | 17:59:14 UTC - in response to Message 3428.

I couldn't manage to make GPUGrid work (always idle dammit) since the CPU request change but it solved the issue by himself last night: WU got into high priority mode and started running. Now I'm on a tight schedule so it keeps working.

But I have to say that since the change my PC have become EXTREMELY MORE SLUGGISH. To the point that I could consider detaching from the project. I also tried lowering the process' nice value but it didn't do any good.

Profile koschi
Avatar
Send message
Joined: 14 Aug 08
Posts: 124
Credit: 466,579,198
RAC: 297,582
Level
Gln
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 3449 - Posted: 28 Oct 2008 | 20:13:42 UTC
Last modified: 28 Oct 2008 | 20:22:12 UTC

Yes, thats one thing I forgot to mention, which bugs me a lot...

On a desktop machine that is actively used, a nice value of 0 for the acemd process is to high. The system feels sluggish...
5 is fine for me, but most of the time I put it back to 19 and only when I leave the computer for some time I switch it to 0.

On my dedicated cruncher, its running at -5, but that one is headless ;-)

If you want to set some nice value to be set automatically, just put the following to roots crontab:

* * * * * renice 19 `ps -ef | grep acemd | grep -v grep | awk '{ print $2 }'`

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 3455 - Posted: 28 Oct 2008 | 20:54:20 UTC - in response to Message 3449.

acemd is using little CPU. It's not the cause of a sluggish system, more that all the cores are busy.

gdf

Profile GDF
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist
Send message
Joined: 14 Mar 07
Posts: 1957
Credit: 629,356
RAC: 0
Level
Gly
Scientific publications
watwatwatwatwat
Message 3456 - Posted: 28 Oct 2008 | 20:54:50 UTC - in response to Message 3455.
Last modified: 28 Oct 2008 | 20:59:01 UTC

Has anybody tried 6.3.19?
gdf

Profile koschi
Avatar
Send message
Joined: 14 Aug 08
Posts: 124
Credit: 466,579,198
RAC: 297,582
Level
Gln
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 3457 - Posted: 28 Oct 2008 | 21:02:04 UTC

I know that it doesn't use much of the CPU, that's not the problem, the problem is the priority...

Low priority (like 19) - desktop is responding quickly and feels smooth
current default priority of 0 - the desktop feels sluggish

Profile Kokomiko
Avatar
Send message
Joined: 18 Jul 08
Posts: 190
Credit: 24,093,690
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 3458 - Posted: 28 Oct 2008 | 21:27:59 UTC

Yesterday I wrote, that my GTX280 on Vista 64 need 34800 sec for one WU. I had reboot the Vista PC and the next WUs are runung with

26676.869 s and
25396.466 s

Looks like a longer uptime makes the calculation slower. Then Vista 64 bit has to reboot often to shorten the calculation time and XP 64 has to boot to clean up the GPU memory.

____________

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 3464 - Posted: 28 Oct 2008 | 22:27:54 UTC - in response to Message 3458.

Looks like a longer uptime makes the calculation slower. Then Vista 64 bit has to reboot often to shorten the calculation time


Are you basing this supposition on one single incident?

MrS
____________
Scanning for our furry friends since Jan 2002

Temujin
Send message
Joined: 12 Jul 07
Posts: 100
Credit: 21,848,502
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwat
Message 3465 - Posted: 28 Oct 2008 | 22:33:39 UTC - in response to Message 3457.

I know that it doesn't use much of the CPU, that's not the problem, the problem is the priority...

Low priority (like 19) - desktop is responding quickly and feels smooth
current default priority of 0 - the desktop feels sluggish

Yep, that works for me

I've added koschi's crontab suggestion
Very sluggish with nice=0, good response with nice=19

Profile Kokomiko
Avatar
Send message
Joined: 18 Jul 08
Posts: 190
Credit: 24,093,690
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 3472 - Posted: 28 Oct 2008 | 23:44:43 UTC - in response to Message 3464.

Looks like a longer uptime makes the calculation slower. Then Vista 64 bit has to reboot often to shorten the calculation time


Are you basing this supposition on one single incident?

MrS


No, 2 in a row. But I will watch this further ...

It's only for the GTX280, our both (Cebion and my) GTX260 running constantly with better times than the GTX280 with a long uptime of the PC. After the reboot the next 2 WUs are faster, then the long run time starts again.

http://www.ps3grid.net/results.php?hostid=7785



____________

Jayargh
Send message
Joined: 21 Dec 07
Posts: 47
Credit: 5,252,135
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwat
Message 3477 - Posted: 29 Oct 2008 | 0:55:32 UTC
Last modified: 29 Oct 2008 | 0:57:03 UTC

I have been getting consistent 26ms and 22000-23000 total runtimes out of my GTX260 EVGA card in Linux Hardy 8.04 latest 2.6.24.21 Kernel.Default is Nice 19 priority btw. My runtimes seem a bit shorter than most other 260's out there that post their numbers.(216 shaders) I have not seen any slowdowns or longer tasks.For me the optimizations so far keep making the tasks run shorter and shorter :)

Profile Venturini Dario[VENETO]
Send message
Joined: 26 Jul 08
Posts: 44
Credit: 4,832,360
RAC: 0
Level
Ala
Scientific publications
watwatwatwatwatwatwatwat
Message 3753 - Posted: 8 Nov 2008 | 0:24:11 UTC - in response to Message 3449.

Yes, thats one thing I forgot to mention, which bugs me a lot...

On a desktop machine that is actively used, a nice value of 0 for the acemd process is to high. The system feels sluggish...
5 is fine for me, but most of the time I put it back to 19 and only when I leave the computer for some time I switch it to 0.

On my dedicated cruncher, its running at -5, but that one is headless ;-)

If you want to set some nice value to be set automatically, just put the following to roots crontab:

* * * * * renice 19 `ps -ef | grep acemd | grep -v grep | awk '{ print $2 }'`



How?

Profile koschi
Avatar
Send message
Joined: 14 Aug 08
Posts: 124
Credit: 466,579,198
RAC: 297,582
Level
Gln
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 3767 - Posted: 8 Nov 2008 | 13:26:20 UTC

you need to become root, or use sudo to start the crontab command as root...
As user you call sudo crontab -e, then it depends on which editor is the default one on your system. Some use pico or nano, which are almost self explaining, as they have the key combinations listet at the bottom. Other distribution use vi, then you should look for a vi tutorial or set the editor variable to a more simple one.

Dont edit roots crontab file (should be somewhere in /var) directly, its better to use crontab because its doing some checking on what you entered and warns you if something is missing...

Post to thread

Message boards : Graphics cards (GPUs) : CPU request is now

//