21) Message boards : Number crunching : LHC on GPU ? (Message 26956)
Posted 7 Nov 2014 by Profile Tom95134
Post:
I am using my GPU to run SETI (cuda50) Tasks. These seem to be well suited for a GPU as they are relatively short in duration (usually about 20 min.) so when a Task grabs the GPU it runs until done. As far as I know, BOINC has no way of managing/doing Task switching on a GPU (but I may be wrong as it has been a long time since I've looked at running other Projects with long GPU Tasks). The last time I tried running a Project with long Tasks on a GPU the Project hogged the GPU.

Now if LHC is going to be pushing a lot more Tasks out to the "crunchers" then maybe it would be worth being able to run them on GPU.

My preference would be to see the effort put into running LHC Tasks on a Virtual machine.

(Just my opinion.)

Tom
22) Message boards : News : Heavy I/O on Windows WUs (Message 26947)
Posted 4 Nov 2014 by Profile Tom95134
Post:
Hi Eric,

I'm not complaining. Just thought you would find the info helpful.

Tom
23) Message boards : News : Heavy I/O on Windows WUs (Message 26945)
Posted 3 Nov 2014 by Profile Tom95134
Post:
I just happened to take a look at the amount of disk space being used by LHC Tasks I have currently running. It is 485+ MB.

Currently 5 Tasks are running.
24) Message boards : News : Heavy I/O on Windows WUs (Message 26944)
Posted 3 Nov 2014 by Profile Tom95134
Post:
Eric,

FYI, I have run a number of the "w-" Tasks and while they take a lot of CPU cycles I have had no problems. I was initially concerned about the long run times but since you calmed me down I don't worry anymore and just let thing crunch away. Also, I do shut the system down for a few hours overnight and then on occasion during the day.

Since the machine I'm using was one configured for a lot of graphics and documentation work it has a large RAM (16GB) and I have allocated 100GB of HDD to BOINC. Now its use is just casual surfing and BOINC. The largest user BOINC of HDD is vLHC @ 5.2GB, 2nd largest is SETI. These two are fairly consistent in their disk usage.

System configuration is:
windows7 Professional x64 - SP1, INTEL i7-2600 3.40GHz, 16GB RAM, NVIDIA GeForce GTS 450, 1TB HDD (c:)

My system is 9955583

Also running SETI, vLHC, & Enigma on the system.

I was running vLHC 4.3.10 until last night and just updated it to 4.3.12.

Tom
25) Message boards : Number crunching : New WUs writing too much to disk (Message 26938)
Posted 31 Oct 2014 by Profile Tom95134
Post:
At one time I did "real work" (business) on my Windowz system. Mainly using Photoshop, Illustrator, and Framemaker. I put these applications in the BOINC->Preferences->Exclusive applications tab. This seemed to work very well for me. Now I don't do a lot of "heavy lifting" so the slowdown is something I can live with.

The only thing I've noticed is the Exclusive applications option may not suspend vLHC crunching. (But I haven't really looked closely.)
26) Message boards : News : 17:00 CET, 15th October, Service back to "normal". (Message 26912)
Posted 16 Oct 2014 by Profile Tom95134
Post:
The Server shows over 500k Tasks ready to send. Are these new Tasks or old work being reshuffled through testing the changes?
27) Message boards : Number crunching : can't open server log file (Message 26869)
Posted 11 Oct 2014 by Profile Tom95134
Post:
Same problem here.

Error message is:
10/11/2014 11:12:15 | LHC@home 1.0 | Server can't open log file (../log_boinc05/scheduler.log)
28) Message boards : Number crunching : LARGE size Work Unit (Message 26861)
Posted 10 Oct 2014 by Profile Tom95134
Post:
Just curious...

I've been crunching the following WU for 8+ hours of CPU time
sd_HL_15_540_1.0_6D_cc_err_Q5__1__s__62.31_60.32__8_10__6__85_1_sixvf_boinc85_1

WORK Unit

This is the most I have experienced in a LHC WU and I was just wondering if there is something "broken".

WU Properties show an Estimated Computation Size of 180000 GFLOPS.

Tom
29) Message boards : News : Three Problems, 22nd May. (Message 26788)
Posted 3 Oct 2014 by Profile Tom95134
Post:
w-b3 Tasks are a disaster. All of the ones I've tried to run crash with Disk Space Error which makes no sense on my system because of the large amount of disk space I allocate to BOINC.

This is a lot of wasted CPU cycles and computing time.
30) Message boards : Number crunching : Error while Computing (Message 26766)
Posted 2 Oct 2014 by Profile Tom95134
Post:
I am starting to see a fairly large number of Error while Computing task failures.

HERE is one.

The failure kicks off the Windows debugging process as you can see in the linked results.

Note: To maximize the number of LHC Tasks processed I suspend other projects I am running with the exception of VirtualLHC@Home.

Tom
31) Message boards : Number crunching : T4T@Home Migration to vLHC@Home (Message 26657)
Posted 13 Jul 2014 by Profile Tom95134
Post:
It was recently announced that T4T@Home is migrating to a new server accompanied by a name change to vLHC@Home. In the thread discussing this migration mention was made that all CERN work is going to use the Oracle VM VirtualBox in the future.

Will LHC@Home also migrate to the VirtualBox environment at some point?

Tom
32) Message boards : LHC@home Science : New data available? (Message 26656)
Posted 13 Jul 2014 by Profile Tom95134
Post:
I'll get Miriam to post some info about the studies.
The new versions of SixTrack I will post shortly
under Status and Plans. Priority is to try and
fix problems with elatest build especially on
Windows. Eric.

Eric,

FYI,
Running Windows 7 Professional (x64), INTEL i7-2600 3.4GHz, 16GB RAM, NVIDIA GeForce GTS450, BOINC 7.2.42 (x64) wxWidgets 2.8.10, VirtualBox 5.3.10.

Not seeing any problems here. All Tasks running to completion although they do run very long. Typical Estimated Time is 2.4 hours but actual running time is 9+ hours. I also run SETI@Home, Enigma@Home and T4T@Home. Even if I suspend these other Projects the LHC@Home Task Running time is 8~9+ hours. You mentioned that this is not a problem but included this info in case it is significant.

Tom
33) Message boards : LHC@home Science : New data available? (Message 26629)
Posted 6 Jul 2014 by Profile Tom95134
Post:
I wonder if these are new WUs or if they are "recycled" to validate a change to sixtrack. They seem to have a much longer running time (~6 hours) than what is typical. All that I have processed have completed without error.
34) Message boards : Number crunching : Long Running Tasks (Message 26602)
Posted 24 Jun 2014 by Profile Tom95134
Post:
Hi Eric,

FYI, all three have now completed and provided Validation for the Tasks.
35) Message boards : Number crunching : Long Running Tasks (Message 26601)
Posted 24 Jun 2014 by Profile Tom95134
Post:
Thank you.

One completed last evening and the other two should complete in about 30 minutes. Although these last two have passed their BOINC Manager deadline but are still okay when I look at the deadline for the Taska in the Task list on my account..
36) Message boards : Number crunching : Long Running Tasks (Message 26599)
Posted 24 Jun 2014 by Profile Tom95134
Post:
I have three Tasks that have been running for 8+ hours.

http://lhcathomeclassic.cern.ch/sixtrack/result.php?resultid=38804767
http://lhcathomeclassic.cern.ch/sixtrack/result.php?resultid=38804793
http://lhcathomeclassic.cern.ch/sixtrack/result.php?resultid=38804797

I have a feeling that there is something "unhealthy" going on with these Tasks. Any ideas from one of the Project Gurus?

Elapsed time keeps decrementing and the Progress (%) seems to be okay but I've never had a task that has run this long.[/quote]
37) Message boards : Number crunching : LHC on GPU ? (Message 26598)
Posted 22 Jun 2014 by Profile Tom95134
Post:
Does the Project have enough work (WU) to justify the effort to develop GPU for LHC?

Are the people at CERN pushing for a faster turn-around of the work they send to LHC?
38) Message boards : Cafe LHC : Computing Preferences (Message 26586)
Posted 2 Jun 2014 by Profile Tom95134
Post:
Leave tasks in memory while suspended: I think it's only useful for buggy projects that crash if you pause them.

For the rest, it depend more on what you do on your pc besides BOINC.
With Intel(R) Core(TM)2 Quad CPU Q9550 @ 2.83GHz & Windows XP SP3, I use at most 75% of processors (3/4) and at most 50% of cpu time, and I still put a good fight against other people regarding working time and cpu time (though I don't follow hardware evolution, so I don't know if they indeed have more recent hardware)

Your suggestion may be fine if you only run BOINC tasks but if you are doing "real work" with your system you may not want to be sacrificing memory while you are doing some Photoshop, Illustrator, Framemaker, etc task.

I have generally had no problems allowing BOINC to dump a Task out of memory when doing other work. It seems to work quite well and I just set those applications that I want to have run without BOINC eating up CPU and Memory into the "exclusive application" tab under Preferences.

I'm not sure if this setting has any effect if you are running Tasks under the Oracle VM as T4T does. At one time it didn't because BOINC didn't control VM but now it might since BOINC Suspends VM quite nicely.
39) Message boards : Number crunching : Status 10th May, Version 451.07 (Message 26451)
Posted 15 May 2014 by Profile Tom95134
Post:
Suspended all other Projects to get LHC work. First Task running now.

Windows 7 x64, INTEL i7


Crunch.. Crunch.. Crunch..

Got a bunch.
40) Message boards : News : LHC@home is back (Message 26450)
Posted 15 May 2014 by Profile Tom95134
Post:
Multiple Gold Stars for Eric.

This wipes out all those previous "Aw Shits".



Previous 20 · Next 20


©2024 CERN