With le nouveau Xeon CPUs supporting Hyper-threading, what is le current wisdom avec regard to using it (or not) on a Hyper-V host machine?
J'étais originally under le impression that turning it on in a virtual host environment could be detrimental as le 'extra' CPUs were pas true cores. However J'ai aussi read (unconfirmed) comments along le lines of MS doing certains hard work to get Hyper-V running well in a Hyper-threading environment.
Does anyone have tout solid information ou experience in this regard? Cheers!
Hyper-V et Hyper-threading : activé ou désactivé ?
Re: Hyper-V et Hyper-threading : activé ou désactivé ?
The old problem avec Hyper-Threading in Virtual Server 2005, sans getting overly technical, is le CPU cache was being poisoned, i.e. it was caching almost nothing parce que le contexts of what was happening on chaque thread were unrelated - causing them to compete for le on-chip cache.
Newer chips have larger et smarter caches so this is less of a problem.
Is it ideal to have on ou off? That really depends on workload. If les deux threads happen to be running le même VM et le même task, alors it would almost certainly be a BIG advantage to have on. If they were doing unrelated things avec a lot of random RAM I/O (several différent VMs par exemple) it would cause seulement half le chip's cache to be disponible to chaque - qui in theory could be slower - In reality it rarely is anymore.
If you have older generation chips bien que you may want to vérifiez le chip cache sizes: In virtualization le larger cache you can get le better. RAM truly is MUCH slower than CPUs - juste pas NEAR as bad as disk drives.
NOTE: What you are reading that says "turn off" was found regarding le chips that were single core avec Hyper-Threading
Newer chips have larger et smarter caches so this is less of a problem.
Is it ideal to have on ou off? That really depends on workload. If les deux threads happen to be running le même VM et le même task, alors it would almost certainly be a BIG advantage to have on. If they were doing unrelated things avec a lot of random RAM I/O (several différent VMs par exemple) it would cause seulement half le chip's cache to be disponible to chaque - qui in theory could be slower - In reality it rarely is anymore.
If you have older generation chips bien que you may want to vérifiez le chip cache sizes: In virtualization le larger cache you can get le better. RAM truly is MUCH slower than CPUs - juste pas NEAR as bad as disk drives.
NOTE: What you are reading that says "turn off" was found regarding le chips that were single core avec Hyper-Threading