2 ms·
Well, this is the reason I hate HyperThreading, does your app consume 50% or 100% - with hyperthreading you have no clue. And that is per core, it becomes incr
by tjoff 10y ago
Well, this is the reason I hate HyperThreading, does your app consume 50% or 100% - with hyperthreading you have no clue.
And that is per core, it becomes increasingly meaningless on a dualcore and on a quadcore and above you might as well replace it with MS Clippy.
And this is before discussing what that percentage really means.
edit: I'm interpreting the downvotes that people are in denial about this ;)
- sqeaky 10y agoDifferent task monitors make this clear in their docs and they tend to stick with it. Read up on the tools you use once and this source of mystery will be gone forever. As far as hyperthreading goes, not understanding some is a bad reason to dislike it. For most developer purposes, pretend it is another CPU until you learn enough about profiling multithreaded apps that you need to care about the difference.
- tjoff 10y agoMy comment was from a user perspective. I don't mind the performance increase from HT (and that is why I still have it enabled). But, it really bothers me that it gets much harder to assess both the total load and the load produced by a single process. I haven't found any useful info on how to deal with this in simple overview user-level tools such as top or the windows task manager. What are the disadvantages? CPU time accounting measurements (sys/usr/idle) as reported by standard tools do not reflect the side-effects of resource sharing between hardware threads It is impossible to correctly measure idle and extrapolate available computing resources https://blogs.oracle.com/partnertech/cpu-utilization-of-multi-threaded-architectures-explained https://blogs.oracle.com/partnertech/cpu-utilization-of-mult...