Advanced search

Message boards : News : New systems in long queue

Author Message
noelia
Send message
Joined: 5 Jul 12
Posts: 35
Credit: 393,375
RAC: 0
Level

Scientific publications
wat
Message 27062 - Posted: 11 Oct 2012 | 15:26:01 UTC

Hi guys,

Just sent a good amount of WUs to the long queue. Credits will vary from 60000 to 70000. Thank all for your time!

Noelia
____________

Profile skgiven
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,995,359,260
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27063 - Posted: 12 Oct 2012 | 13:04:56 UTC - in response to Message 27062.

New research?
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

noelia
Send message
Joined: 5 Jul 12
Posts: 35
Credit: 393,375
RAC: 0
Level

Scientific publications
wat
Message 27064 - Posted: 12 Oct 2012 | 16:09:35 UTC - in response to Message 27063.

It is actually the continuation of a system I sent last month (domain SH2), and worked quite well, but I need more data :)

Simba123
Send message
Joined: 5 Dec 11
Posts: 147
Credit: 69,970,684
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 27065 - Posted: 13 Oct 2012 | 10:09:59 UTC

and we'll be happy to oblige :)


Profile ritterm
Avatar
Send message
Joined: 31 Jul 09
Posts: 88
Credit: 244,413,897
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27066 - Posted: 14 Oct 2012 | 11:39:49 UTC - in response to Message 27062.
Last modified: 14 Oct 2012 | 11:40:11 UTC

Just sent a good amount of WUs to the long queue...

I've been fishing for these for nearly 3 days and have come up empty...<sniff, sniff> :'(
____________

TheFiend
Send message
Joined: 26 Aug 11
Posts: 99
Credit: 2,500,112,138
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27080 - Posted: 16 Oct 2012 | 18:56:27 UTC - in response to Message 27066.

Just sent a good amount of WUs to the long queue...

I've been fishing for these for nearly 3 days and have come up empty...<sniff, sniff> :'(


Just caught my first one...... looks to be around 8 hours on my GTX670

Profile ritterm
Avatar
Send message
Joined: 31 Jul 09
Posts: 88
Credit: 244,413,897
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27082 - Posted: 17 Oct 2012 | 2:31:16 UTC - in response to Message 27080.

Just caught my first one...... looks to be around 8 hours on my GTX670

I finally caught one for my GTX 570. It's in line behind a NATHAN that won't finish for about 8 hours, though.
____________

Rayzor
Send message
Joined: 19 Jan 11
Posts: 13
Credit: 294,225,579
RAC: 0
Level
Asn
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 27083 - Posted: 17 Oct 2012 | 6:11:16 UTC
Last modified: 17 Oct 2012 | 6:49:48 UTC

Caught one too, (ala_structure_191-NOELIA_1CPU_sh2fragment-0-3-RND2901)

Should take about 10.5 hours, on a GTX 670

This WU is using 440mb memory, with 71% GPU usage, Windows 7 Ultimate

TheFiend
Send message
Joined: 26 Aug 11
Posts: 99
Credit: 2,500,112,138
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27084 - Posted: 17 Oct 2012 | 7:10:50 UTC - in response to Message 27080.

Just sent a good amount of WUs to the long queue...

I've been fishing for these for nearly 3 days and have come up empty...<sniff, sniff> :'(


Just caught my first one...... looks to be around 8 hours on my GTX670


Finished in 8.6 hours on my GTX670 and scored 114300, also used less CPU time!!!

http://www.gpugrid.net/result.php?resultid=5923417

Profile ritterm
Avatar
Send message
Joined: 31 Jul 09
Posts: 88
Credit: 244,413,897
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27088 - Posted: 18 Oct 2012 | 0:31:48 UTC - in response to Message 27084.

Finished in 8.6 hours on my GTX670 and scored 114300...

About 11.6 hours for this one, 3761726, on my stock GTX 570. Good for 95,250. Crunching my second one now. :-)
____________

Zac
Send message
Joined: 24 Oct 12
Posts: 1
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 27151 - Posted: 24 Oct 2012 | 22:38:15 UTC

Hi there!
I'm crunching with the NVS 4200 of my laptop. It's not the newest GPU but I'd like to help...
I received one of this WU I think, the problem is that it says the remaining time is about 120h, so I'll not be able to return the WU before the deadline.
Is there anything I can do?

Thanks!

Profile Carlesa25
Avatar
Send message
Joined: 13 Nov 10
Posts: 328
Credit: 72,619,453
RAC: 0
Level
Thr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27154 - Posted: 25 Oct 2012 | 10:18:09 UTC - in response to Message 27151.
Last modified: 25 Oct 2012 | 10:19:07 UTC

Hello Zac: I think it best to cancel the task and downloads only short drives - ACEMD standard-which are also very valuable work.

GPUGRID requires high-level GPUs. Greetings.

Granite T. Rock
Send message
Joined: 7 Sep 09
Posts: 3
Credit: 7,778,853
RAC: 0
Level
Ser
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 27155 - Posted: 25 Oct 2012 | 14:37:17 UTC - in response to Message 27154.

Generally I thought BOINC downloaded work units based upon if it thinks you can finish it on time? Or is that something implemented server side? I only take a general interest in my work units once every couple of months. Hate to think people's cycles are being wasted because the clients are downloading more than they can chew.

Profile skgiven
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,995,359,260
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27165 - Posted: 27 Oct 2012 | 12:56:39 UTC - in response to Message 27151.
Last modified: 27 Oct 2012 | 12:58:46 UTC

The Quadro NVS 4200M (GF119) only has 48 cuda cores (shaders), of which only 32 would be used (CC2.1). It's also GDDR3 and only 155GFlops peak, so I doubt that it would finish any task in time. I would not recommend using that GPU here. If you still want to use it I suggest you try another project (it might work on POEM).

PS. You get work so long as your GPU's Compute Capability is still accepted. So you won't get tasks on a CC1.1 card here, and for some projects CC1.2 is no longer supported. It's based on 'can it run the tasks' rather than how 'long they will take'/performance.
You might think it would be useful if a list of excluded cards was generated server side, based upon the inability to complete and report tasks on time. However, some cards work on one setup but not another, some cards are overclocked, some underclocked, some are DDR3, some are DDR5, some are on PCIE 1.1 boards and others on PCIE2 or PCIE3 boards, some are supported by fast processors and fast RAM, others not... So you can't easily stereotype all the cards, to work out if they are useful in advance.
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

[AF>Belgique] bill1170
Send message
Joined: 4 Jan 09
Posts: 13
Credit: 1,195,187,199
RAC: 5,433,293
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27670 - Posted: 12 Dec 2012 | 14:07:04 UTC - in response to Message 27165.

Dear Noelia,

You may not have seen the topic in the "number crunching" section related to relatively frequent problems with some of your "NOELIA" wus presently running in the long queue.

Could you please advise ?

Thanks in advance.

noelia
Send message
Joined: 5 Jul 12
Posts: 35
Credit: 393,375
RAC: 0
Level

Scientific publications
wat
Message 27672 - Posted: 12 Dec 2012 | 15:19:32 UTC

Dear volunteers,

Batch group "sh2forces" requires extensive CPU usage and that's why some of you have noticed a big slow down. Same problem as we were having with previous systems:
http://www.gpugrid.org/forum_thread.php?id=3116 but we won't send any of this kind anymore.
The group had 100 WUs, but it is aborted now.

Noelia

jlhal
Send message
Joined: 1 Mar 10
Posts: 147
Credit: 1,077,535,540
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27674 - Posted: 12 Dec 2012 | 17:00:39 UTC - in response to Message 27672.

Dear volunteers,

Batch group "sh2forces" requires extensive CPU usage and that's why some of you have noticed a big slow down. Same problem as we were having with previous systems:
http://www.gpugrid.org/forum_thread.php?id=3116 but we won't send any of this kind anymore.
The group had 100 WUs, but it is aborted now.

Noelia


Thanks a lot !
____________
Lubuntu 16.04.1 LTS x64

Profile skgiven
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,995,359,260
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27682 - Posted: 13 Dec 2012 | 0:05:49 UTC - in response to Message 27672.

Dear volunteers,

Batch group "sh2forces" requires extensive CPU usage and that's why some of you have noticed a big slow down. Same problem as we were having with previous systems:
http://www.gpugrid.org/forum_thread.php?id=3116 but we won't send any of this kind anymore.
The group had 100 WUs, but it is aborted now.

Noelia

Does this mean we should abort these tasks?

I was sent two sh2forces tasks yesterday evening/overnight. They are now at >65% and 92%. Just under 2h left on one!

Anyway, I did some testing on them. To cut a long story short, the performance is directly limited by the tasks CPU requirement. A solution could be going multi-threaded (mt), running some of those CPU cycles on the GPU (if possible), or running multiple WU's on the same GPU.

Beta testing would be a good idea too!

The slowness isn't due to PCIE bandwidth, I tested at PCIE3.0 x16 and PCIE2.0 x8; no difference in performance.
It's also not a controller bottleneck, or too much memory being used.
It's just down to the CPU usage and CPU to GPU performance ratio. A single thread from a fast CPU isn't capable of supporting the requirements of a low to mid range GPU. Even a GTX460 supported by a highly overclocked i7 would not yield good GPU utilization:

By crunching fewer CPU tasks I was able to improve performance, somewhat.
(i7-3770K@4.2GHz HT on, 8GB DDR3-2133, SATA6 SSD, W7x64)

When CPU usage was at 98% (crunching CPU tasks on 6 threads + 2 GPU tasks)
GPU Load - GTX660Ti 7% to 10% - GTX470 10%
Mem Ctrl - GTX660Ti 3% - GTX470 2%

When CPU usage was at 46% (crunching CPU tasks on 2 threads + 2 GPU tasks)
GPU Load - GTX660Ti 9% to 12% - GTX470 11% to 15%
Mem Ctrl - GTX660Ti 4% - GTX470 3%

When CPU usage was at 21% (only crunching 2 GPU tasks)
GPU Load - GTX660Ti 12% to 14% - GTX470 15% to 17%
Mem Ctrl - GTX660Ti 5% - GTX470 3%

With HT off and CPU usage was at 43% (only crunching 2 GPU tasks)
GPU Load - GTX660Ti 13% - GTX470 17%
Mem Ctrl - GTX660Ti 5% - GTX470 4%

The GPU temperatures and fan speeds rose after freeing CPU cores and turning HT off (not that I would normally recommend turning it off). Even with such optimizations however the GTX660Ti was only using 35% of it's power consumption capabilities.
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2343
Credit: 16,201,255,749
RAC: 0
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27683 - Posted: 13 Dec 2012 | 0:43:52 UTC - in response to Message 27672.

Batch group "sh2forces" requires extensive CPU usage and that's why some of you have noticed a big slow down. Same problem as we were having with previous systems:
http://www.gpugrid.org/forum_thread.php?id=3116 but we won't send any of this kind anymore.
The group had 100 WUs, but it is aborted now.

Strange.
Very strange.
You didn't said who aborted these workunits...
It seems that you didn't successfully aborted this batch, so the crunchers have do it one by one.

klepel
Send message
Joined: 23 Dec 09
Posts: 189
Credit: 4,631,406,793
RAC: 2,310,921
Level
Arg
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27694 - Posted: 13 Dec 2012 | 23:00:58 UTC - in response to Message 27683.

Noelia, If these work unites are important for your work I will crunch all 100 WUs "sh2forces" happily. Isn’t it about science? As I have written before, yes credits are important for me as well, but if everybody gets one or two of these slow not optimal WUs to crunch, it affects all and nobody gets really hurt credit wise.

HA-SOFT, s.r.o.
Send message
Joined: 3 Oct 11
Posts: 100
Credit: 5,879,292,399
RAC: 0
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27696 - Posted: 14 Dec 2012 | 11:28:38 UTC - in response to Message 27694.

No problem for me also.

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2343
Credit: 16,201,255,749
RAC: 0
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27697 - Posted: 14 Dec 2012 | 12:00:48 UTC - in response to Message 27696.

No problem for me also.

The performance of these workunits on Linux is ~3 times better than on Windows.
There is no much difference between XP or 7 regarding these workunits.
I've aborted another one, it was 12% complete after 5 hours. It's not so popular.
I've suspended every CPU task to see if there's any improvement in GPU usage, and my GPU usage's gone up by 3% to 11%.
I still consider these workunits as a waste of resources (on Windows).

Profile skgiven
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,995,359,260
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27698 - Posted: 14 Dec 2012 | 13:05:13 UTC - in response to Message 27697.
Last modified: 14 Dec 2012 | 13:06:28 UTC

No problem for me also.

The performance of these workunits on Linux is ~3 times better than on Windows.


Linux GTX 680:
Run time 44,471.89
CPU time 38,477.50

W7 (GTX660Ti, GTX470):

ala_structure_881-NOELIA_sh2forces-3-5-RND6660_1 3934692 139265 12 Dec 2012 | 0:34:09 UTC 13 Dec 2012 | 7:08:28 UTC Completed and validated 86,697.19 78,124.14 79,375.00 Long runs (8-12 hours on fastest card) v6.16 (cuda42)
ala_structure_241-NOELIA_sh2forces-1-5-RND3374_3 3931355 139265 11 Dec 2012 | 19:34:19 UTC 13 Dec 2012 | 1:47:29 UTC Completed and validated 88,036.52 79,443.15 79,375.00 Long runs (8-12 hours on fastest card) v6.16 (cuda42)

I don't know what else (HA-SOFT, s.r.o.)'s Linux system was doing, if anything, (possibly some WCG or Rosetta CPU tasks) but if it wasn't I would say it's roughly twice as fast as W7. If it was crunching CPU tasks then I would agree it could be close to 3 times as fast. My tasks were initially run while crunching CPU tasks. Around half way through I optimized for the GPUGrid tasks.

The limiting factor on Windows appears to be the CPU requirement, hence the GTX470 was almost as fast as the GTX660Ti.

Even on Linux, and assuming it's 3times as fast, GPU utilization would only be ~39%. If the research is so important, make it Linux only, or better still, make an mt app and feed the GPU properly.
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

HA-SOFT, s.r.o.
Send message
Joined: 3 Oct 11
Posts: 100
Credit: 5,879,292,399
RAC: 0
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27699 - Posted: 14 Dec 2012 | 13:10:35 UTC - in response to Message 27698.


I don't know what else (HA-SOFT, s.r.o.)'s Linux system was doing, if anything, (possibly some WCG or Rosetta CPU tasks) but if it wasn't I would say it's roughly twice as fast as W7. If it was crunching CPU tasks then I would agree it could be close to 3 times as fast. My tasks were initially run while crunching CPU tasks. Around half way through I optimized for the GPUGrid tasks.


6 HT Cores WCG
2 HT Cores left for GPUGRID

Profile skgiven
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,995,359,260
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 27700 - Posted: 14 Dec 2012 | 13:41:56 UTC - in response to Message 27699.

Thanks, ~3times as fast it is then ;)
If the utilization really is ~39% on Linux with 6CPU threads used by WCG, then it might mean it has potential to rise to ~50% by using less CPU cores.

At 50% GPU utilization I would be reasonably happy to run a few of these on Linux, but on a normal Windows crunching setup, with a highly recommended GPU and recommended GPUGRID settings, getting 7% GPU utilization isn't acceptable. If it was the GT640 would be at the top of the recommended GPU list! We don't buy high end GPU's to tick over at 7%.
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Post to thread

Message boards : News : New systems in long queue

//