1) Message boards : Questions and problems : multiple cpu core issue (Message 98407) Posted 9 May 2020 by jdzukley Post: I am running version 7.16.5 And new information: I noticed that boinc CPU execution became "hung". noticecd by viewing task manage, and not seeing any percent increase in task progress. I suspended all tasks, then shut down boinc, restarted, and the tasks that were identified previously as 8 core cpu's including the task in progress, changed to 12 core CPU (as settings were 100% on this 12 core system. I released the in progress task, now labeled as 12 core, and all appears well, running at 95+/-% system capacity essentially using all cores., I appear to not have an issue or problem. Very interesting to see that multi core tasks - at least nbody tasks - can have their multi cores usage adjusted during the boinc start process.

2) Message boards : Questions and problems : multiple cpu core issue (Message 98399) Posted 9 May 2020 by jdzukley Post: I have identified an issue with jobs that use multiple cpu cores. I do like to use Milkyway@home, using the nbody application which uses multiple cores. I would set in preferences to use 70% of the CPU, and request nbody applications. At 70% of CPU with 12 cores, I would obtain tasks that use 8 cores. I would like to use the the reaming cores on other applications. So I would set Milkyway to no new tasks, reset the CPU preferences back to say 95%. Other apps like world community grid or Rosetta would run in these additional CPU cores - and then stop - go into "waiting to run". If I suspend the milkyway 8 core nbody job, the cores will become active again. When I restart the milkyway nbody task, tasks will continue to run in the remaining available CPU cores - for some period of time. I do have 2 GPU cards. It appears to me that whenever a condition arises that causes BOINC to recalcuate the availabe CPU's, that tasks such as nbody using multicore CPU's will somehow "assume" that they will use 100% of the cores (except for GPU tasks).. Bottom line, if I set my CPU available to 100% - 12 cores, and have an nbody task running identified at 8 CPU cores, after a short time, only the 8 CPU core nbody task ( and GPU tasks) are running, all other tasks are waiting to run mode.

3) Message boards : BOINC Manager : Suggestions For New Or Updated Features (Message 59225) Posted 1 Jan 2015 by jdzukley Post: I have a Windows based computer with 12 cores and 2 NVida GPU's. Currently, I set my list of project to only accept tasks from GPUGRID and The Lattice Project. The Lattice Project is currently running short - 1 hour tasks with very quick turnaround required, i.e. 6 hours. My Boinc network settings are 0.02 min and 0.03 additional max. Often the Boinc manager turns off the second GPU even though there are only a few - in this last case 3 work units in the job not started que. Bottom line, There is plenty of time for the currently running jobs to finish, and neither of the GPU's should be turned off

4) Message boards : GPUs : multi GPU's may cause CPU cores to not be "fully" utilized (Message 49215) Posted 19 May 2013 by jdzukley Post: I searched for this topic and apologize if I missed prior discussions... If this was discussed, please advise post link to the discussion thread. I have a i7 12 core machine with 2 GTX 580 cards, 64g memory. Boinc manager version 7.0.64, 2.8.10. I normally set the Bionc manager to 99% of CPU core as that appears to maximize this systems average output for most of the typical tasks executed. I believe output could be further optimized if the CPU usage setting was set to 100% and the Boinc manager determined the number of core CPU's to be: Core CPU allowed less (the rounded up (sum requirements of all GPU CPU's requirements)). It appears to me that the Boinc manager currently rounds each GPU CPU requirement individually (rather then cumulative). So the computer could be over on net CPU execution at the expense of the GPU's or vise versa where GPU's are near 100% utilized and CPU cores are under utilized. I believe this will optimize both GPU and CPU core task execution. Granted this can be done manually for the current work settings, by monitoring system monitors..... Anyway, I look forward to any / all discussion.

---

Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.