You can now assign processes to specific CPUs or groups of CPUs in System Monitor, known as setting CPU affinity. (Taras Oleksyn, KDE Bugzilla #429151)
Huh. I wonder when this is needed to set it after a process is started. Is there an actual use case for this? I understand what this does, but why would anyone change it? Is there a real world example someone could give me?
Some generally older software has buggy multi threading implementation, and forcing its threads to execute “interleaved, but never simultaneously” often improves stability or even completely hides the issue.
Also I know some games used to have big performance issues if they had too many hardware threads available, but I don’t know why exactly
There’s also modern CPU structure, where there might be a performance hit if you involve cores that are attached to a different cache or don’t have direct access to memory. Threadripper ran into that sort of thing.
Some processors, like Ryzen CPUs above 8 cores, have their cores physically split onto multiple chiplets, and having a process split across two or more chiplets can introduce overhead transferring data between them.
Huh. I wonder when this is needed to set it after a process is started. Is there an actual use case for this? I understand what this does, but why would anyone change it? Is there a real world example someone could give me?
Some generally older software has buggy multi threading implementation, and forcing its threads to execute “interleaved, but never simultaneously” often improves stability or even completely hides the issue.
Also I know some games used to have big performance issues if they had too many hardware threads available, but I don’t know why exactly
There’s also modern CPU structure, where there might be a performance hit if you involve cores that are attached to a different cache or don’t have direct access to memory. Threadripper ran into that sort of thing.
Some older games (example: GTA 4) have broken physics with modern CPUs due to core count.
Emulation i would guess?
Some processors, like Ryzen CPUs above 8 cores, have their cores physically split onto multiple chiplets, and having a process split across two or more chiplets can introduce overhead transferring data between them.