Message boards :
Number crunching :
Short runs same duration than long runs !?
Message board moderation
| Author | Message |
|---|---|
Maurice GouloisSend message Joined: 22 Feb 09 Posts: 10 Credit: 103,904,673 RAC: 0 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Hi there, look at that pc, https://www.gpugrid.net/results.php?hostid=217261, there seems to have same duration for all recent workunits whatever "run" they belong. Not a problem but guessing if normal behaviour. (in fact, short ones have been longer than long ones :))
|
dskagcommunitySend message Joined: 28 Apr 11 Posts: 463 Credit: 979,266,958 RAC: 76,910 Level ![]() Scientific publications ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]()
|
Ah and i thought this would be a problem of my at this time unstable running GPUs with the long and short units same runtime. So i seconed that behavour, i saw that too ^^ DSKAG Austria: http://www.dskag.at
|
|
Send message Joined: 1 Feb 14 Posts: 9 Credit: 5,837,925 RAC: 0 Level ![]() Scientific publications ![]() ![]()
|
I have a low end card and only do the short run NOELIA units. I see that as the wu become more complex they take longer to crunch, a few months ago they ranged from 36-42 hours, then went up to 60+ and now 70+ is becoming the norm though a couple were c. 48(cpu time/run time = c. 0.15). Since I was advised not to do long units due to high risk of failures I would appreciate advice whether I should continue or just focus on my current sole alternative of Poem@Home. I have an SSD HD and have just upgraded to W10, if that's relevant. Maybe too the credits should reflect the longer times - I used to have a small chance of getting the 25% but that is no longer possible - not that this would influence my decision to continue - but it would be nice to get to 5m. |
©2026 Universitat Pompeu Fabra