Questions and Answers :
Wish list :
2 performance related wishes
Message board moderation
Author | Message |
---|---|
Rept-Tile Send message Joined: 18 Oct 03 Posts: 20 Credit: 28,144,831 RAC: 0 |
Both are belonged to the situation for offline multi-CPU with HT crunching computers at the moment when tasks are almost finished and the tasks actual number is less than nominal. (Please drive me to the topic link if similar suggesstions have been asked already, thanks in advance). 1. Make an option in client preferences to include or to not include HT-emulated processors while counting a nominal number of simultaneous tasks. Now it's always equal to the number of CPUs (both real and emulated). 2. Make an option in client's behaviour to increase the percentage of using the CPU for avery active thread. For example. I have 4 CPUs and nominal count of simultaneous is 4 too, and I set 100% of summary CPU usage in "Preferences": - While the computer is crunching 4 tasks, the percentage of CPU usage for every application thread is 25%. - When one of 4 tasks is finished (3 remaining and no new tasks by some reason) every thread is still using 25%, it makes 75% of total CPUs' usage in sum. - What do I want in a last case: to have an option to automatically increase a percentage for every thread to 33.333..% to make a summary usage of 100% when the actual number of WUs becomes 3. Thank you. |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
1) There is already an option. You can set the maximum number of CPUs to use. 2) BOINC uses one process per CPU. Windows reports the total of all CPUs as 100%. If you have 4 CPUs and one is idle, windows will report this as 75% usage and 25% idle. The science applications are single threaded and cannot use more than a single CPU. BOINC WIKI |
Rept-Tile Send message Joined: 18 Oct 03 Posts: 20 Credit: 28,144,831 RAC: 0 |
John McLeod VII wrote: 1) There is already an option. You can set the maximum number of CPUs to use. I talk about a little different thing. When I'm cutting down number of CPUs in "Processor preferences", the total load of the system falls down too. I'd like to know about possibility to load all CPUs' resources by a lesser WUs number and, optionally, to have a possibility to count that number automatically, excluding HT CPUs. I understand, it flows directly to the next question below. John McLeod VII wrote: 2) BOINC uses one process per CPU. Windows reports the total of all CPUs as 100%. If you have 4 CPUs and one is idle, windows will report this as 75% usage and 25% idle. The science applications are single threaded and cannot use more than a single CPU. Ah! It must be a stumbling block for all of my questions. Hmm.. they are really single-threaded ? If so, I would like to get some explanation for the next screenshots. 1st and 2nd were taken in 4-CPU (2 real + 2 HT emulated) Xeon system while working with 3 WUs. Performance Processes How can it be if S@H process (or at least it's mostly resource-taking counting part) is single-threaded?? And the 3rd displays 4 process threads in S@H 5.27 application. It's from 1-processor Celeron, so just 1 S@H process is in the memory. S@H threads viewed by Process Explorer I asked the same questions here, in BOINC forum, but without significant or explaining answers until the moment.. Thanks. |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
The computation is indeed single threaded. There are a couple of threads for communication with the daemon that will not use much CPU time. BOINC is currently designed on the assumption that computation will be single threaded as this is how the BOINC CPU scheduler tracks resource usage, one science application process to one CPU slot. There has been discussion of allowing multi threaded applications, but it has not been designed (this is harder than it might seem at first glance). BOINC WIKI |
Rept-Tile Send message Joined: 18 Oct 03 Posts: 20 Credit: 28,144,831 RAC: 0 |
John McLeod VII wrote: There has been discussion of allowing multi threaded applications, but it has not been designed (this is harder than it might seem at first glance). Agree, like with all practical realizations. Is it possible to find that discussion? It's really interesting for me. Thank you for your answers! But I'm still confused by my 1st (in conjunction with 2nd) screenshot. If we have only 3 counting threads while computing 3 WUs on 4 CPUs, what makes PERMANENT load approximately 75% for the 4th CPU?? It's not any of other processes/applications, in shot 2 the list is sorted by "CPU" column. |
Rept-Tile Send message Joined: 18 Oct 03 Posts: 20 Credit: 28,144,831 RAC: 0 |
|
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.