| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: OpenPandemics - COVID-19 Project
Thread: Work unit availability |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 822
|
|
| Author |
|
|
True54Blue
Advanced Cruncher Joined: Nov 17, 2004 Post Count: 97 Status: Offline Project Badges: 
|
I did get more than usual just recently of the 0041___ and 0042____ batches but they've dried up. They all seem to have over 100 jobs in each WU which is nice as it gives the GPU something more substantial to bite down on than kibble :)
---------------------------------------- |
||
|
|
Martin Schnellinger
Advanced Cruncher Joined: Apr 29, 2007 Post Count: 128 Status: Offline Project Badges: 
|
Hello,
maybe it would be a good idea to give the information when this or that batch citation "dried up". Each workunit has a "sent time" Page: 1 2 3 Last > Result Name Device Name Status Sent Time Time Due / Return Time CPU Time / Elapsed Time (hours) Claimed/ Granted BOINC Credit OPN1_0048774_00440_0 ms-PC In Progress 5/20/21 17:55:50 5/27/21 17:55:50 0.00 / 0.00 0.0 / 0.0 Mentioning the sent time would make this huge thread more understandable and trackable for technicians form WCG. Just an humble idea All the best, keep crunching on MS |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1327 Status: Offline Project Badges: 
|
The WUs from the OPNG_0041... batch are really strange. They take a lot longer than all the other ones so far, and most of the time the GPU stays un-utilized. Is this by design? Or is something wrong with them? I noticed this during the stress test also. the vast majority of the runtime difference comes down to the AutoGrid run time. on the older tasks it only takes a few seconds. on these new tasks, AutoGrid takes several minutes. This appears to be a CPU-based setup step before anything runs on the GPU, so the GPU will be idle during this time. so far no one has explained the reason for this. AutoGrid looks at the receptor data file and makes various "grids" used by AutoDock. The receptors for these high-numbered batches are much larger than those usually targeted so AutoGrid takes a lot longer to compute the relevant grids... Most receptor files seem to be about 86.4KB with about 1100 atoms, but the big receptors have file sizes around 2.5MB and nearly 32000 atoms - about 30 times the amount of data! On my fastest system, the smaller receptors seem to need between 3 and 7 seconds for AutoGrid whilst the larger ones seem to need 50 to 60 seconds at the moment. The ones during the stress test were taking 130 seconds or so, but I think that was a rigid receptor, whilst the current ones seem to have a flex file as well. (I have no idea whether that makes that much difference to AutoGrid behaviour, however...) Cheers - Al. P.S. The above is based on observation of data, not examination of source code! [Edited first paragraph to rephrase AutoGrid action.] [Edit 1 times, last edit by alanb1951 at May 20, 2021 9:40:49 PM] |
||
|
|
biini
Senior Cruncher Finland Joined: Jan 25, 2007 Post Count: 334 Status: Offline Project Badges: 
|
I did get more than usual just recently of the 0041___ and 0042____ batches but they've dried up. They all seem to have over 100 jobs in each WU which is nice as it gives the GPU something more substantial to bite down on than kibble :) I noticed this also. Actually my rtx was fully occupied for 2,5hours. Havent seen that lately :-) E: just while I was writing this, got another batch for rtx :-o ---------------------------------------- [Edit 1 times, last edit by biini at May 20, 2021 9:46:24 PM] |
||
|
|
kittyman
Advanced Cruncher Joined: May 14, 2020 Post Count: 140 Status: Offline Project Badges: 
|
Well, the kitties congratulate you on your good fortune. Their kibble bowl has been rather empty for the last 2 days and their RAC is diving.
----------------------------------------Meow  |
||
|
|
zdnko
Senior Cruncher Joined: Dec 1, 2005 Post Count: 235 Status: Offline Project Badges: 
|
I got two errors "exceeded elapsed time limit 9454.77"
OPNG_ 0041873_ 00044 OPNG_ 0041873_ 00003 |
||
|
|
Crystal Pellet
Veteran Cruncher Joined: May 21, 2008 Post Count: 1411 Status: Offline Project Badges: 
|
I got two errors "exceeded elapsed time limit 9454.77" OPNG_ 0041873_ 00044 OPNG_ 0041873_ 00003 I suppose, this was on an Intel iGPU? |
||
|
|
Lasantha Bandara
Cruncher Joined: Jul 9, 2010 Post Count: 3 Status: Offline Project Badges:
|
I havent received WU in a while is this normal?My gpu is zero work load for days now
 |
||
|
|
squid
Advanced Cruncher Germany Joined: May 15, 2020 Post Count: 56 Status: Offline Project Badges: 
|
Low-performing GPUs like iGPU get WUs on a regular basis. GPUs with a lot of power like RTX 3070 get WUs every now and then.
|
||
|
|
Acibant
Advanced Cruncher USA Joined: Apr 15, 2020 Post Count: 126 Status: Offline Project Badges:
|
I havent received WU in a while is this normal?My gpu is zero work load for days now GPUGRID uses the BOINC client and you can connect to that project using the same client used for World Community Grid (which is just a branded BOINC client). In your preferences on that site you can set the resource share to 0, making it a backup project. Whenever WCG has GPU work units they will get priority since you would have a "project weight" in your device profiles for WCG set to more than 0.  |
||
|
|
|