| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
| World Community Grid Forums
|
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 3
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
hi everybody,
i am a little bit confused about this. how is this average calculated in the first place? is it an average over 1 day? over a week? a month? i have 6 device installations, and they all run jobs 24/7, however my average runtime per calendar day is like 5 days and 8 hours. how does how busy your CPU is affect the calculated runtime. i know that it would slow jobs down so you get less points or jobs done over time, but 1 hour of BOINC working on a job should equal 1 hour of runtime anyway? i also have a question about how hyper-threading your CPU affetcts the calculated runtime. does it double it? it would seem to me like it should since for every hour of time i have 2 hours of job done. thank you! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello potsik,
I cannot answer most of your questions, but I can say that BOINC keeps track of time run much more carefully than UD. The UD client just tracks run time and does not shrink it by the percent utilized. Hyperthreading can gain you something, but a second thread increases throughput only (on average) about 15% of the time. And on our scientific applications, both competing for the same FPU, the second thread might actually run much less than that. Lawrence |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
potsik,
The BOINC agent only counts the time the process is actually running as run time (this is as opposed to wall clock time). You appear to be running at a little less than 97% run time vs. wall clock time right now. Background processes on your computer(s) will probably take up the that remaining 3%, so I'd say you're operating at about as efficient a rate as you can get. |
||
|
|
|