Message boards : Questions and problems : Why is 60% CPU time and 100% CPU threads the default?
Message board moderation
Author | Message |
---|---|
Send message Joined: 18 Apr 20 Posts: 3 ![]() |
I'm wondering what's the reason behind choosing this setting (in the title) rather than the opposite (100% time and 60% threads). - Is it a better way to prevent high CPU temperatures? - Is it a better way to make room for other applications? - Is it only to allow low-end CPUs (e.g. with 4 threads) to make room for other applications without removing too much resources ? - if I had 100 threads, would this setting or the opposite one make any difference in temperature or performance ? |
Send message Joined: 10 Apr 20 Posts: 9 ![]() |
There are single-core CPUs, and setting threads count to 60% would make them not running anything. |
Send message Joined: 18 Apr 20 Posts: 3 ![]() |
Thanks. So that would be a yes to my 3rd guess in the list. Is that the sole reason though? What about my other guesses? |
Send message Joined: 8 Nov 19 Posts: 718 ![]() |
There are single-core CPUs, and setting threads count to 60% would make them not running anything. What PC you can buy today, still has a single core CPU? Even if they had, they would be with hyperthreading, and run 1 CPU thread. The last truly 1CPU core CPU I remember coming out, was in the Pentium 4 era. That's the late 2000s, or almost 20 years ago. I sincerely hope no one is crunching on that, as a $100 netbook will do faster! |
![]() Send message Joined: 28 Jun 10 Posts: 2842 ![]() |
There are single-core CPUs, and setting threads count to 60% would make them not running anything. Don't know about other projects but Pentium4 has SSE instruction set but not SSE2 and a few years ago now all CPDN tasks started requiring SSE2 as a minimum in order to run tasks. I suspect many if not most projects now require that so perhaps it is at least time to discuss whether the defaults should change, remembering that the defaults have to be relatively non intrusive to the #setandforget crowd. |
Copyright © 2025 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.