Message boards :
Number crunching :
Run Time Behaviour
Message board moderation
Author | Message |
---|---|
Send message Joined: 14 Dec 09 Posts: 161 Credit: 589,318,064 RAC: 0 |
Hello, http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=31031387 When you check this wu, the completion time seems very low. Such behaviour occurred when i restart the rig during crunching and although computation continued from where it left, the runtime calculation restarted. Shouldn't the runtime also continue from where it left? Regards. |
Send message Joined: 18 Jul 09 Posts: 300 Credit: 303,562,776 RAC: 0 |
If the total run time was what you expect for your card, I would not be concerned. The part I find interesting is that computation continued from where it left off, I cannot do that. But with 90 second WUs I don't worry about that either. |
Send message Joined: 8 Feb 08 Posts: 261 Credit: 104,050,322 RAC: 0 |
Checkpointing came with separation v0.80 but depends on your BOINC settings. Might be a bug not picking up the previous runtime. Nothing serious. With those short runtimes on modern gpus not many people will even see a checkpoint happening because their checkpoint interval is far longer than a WU runs. |
Send message Joined: 18 Jul 09 Posts: 300 Credit: 303,562,776 RAC: 0 |
Checkpointing came with separation v0.80 but depends on your BOINC settings. So checkpointing is based upon time the WU has run rather than a milestone? Is this true with all projects? |
©2024 Astroinformatics Group