Core parking happened. SMT requires OS scheduler to be aware of what cores are not real.But I kept reading that RyZen SMT was considerably better than Intel's HT?
Then how did we get the earlier findings that it was better? Did they not use Windows to test?Core parking happened. SMT requires OS scheduler to be aware of what cores are not real.
AMD stock price has dropped 2% since 9 am.
Waiting for someone to do a proper 4K gaming review.
Ryzen have HT bug in games -10-15% performance drop.
http://www.hardware.fr/articles/956-7/impact-smt-ht.html
It probably only matters for games.Then how did we get the earlier findings that it was better? Did they not use Windows to test?
Ryzen have HT bug in games -10-15% performance drop.
http://www.hardware.fr/articles/956-7/impact-smt-ht.html
Ryzen have HT bug in games -10-15% performance drop.
http://www.hardware.fr/articles/956-7/impact-smt-ht.html
They used workloads that take up all the threads available. No core parking issues with these.Then how did we get the earlier findings that it was better? Did they not use Windows to test?
They will be the same on G4560 and 1800X, because GPU bottlenecks, man.Waiting for someone to do a proper 4K gaming review.
Not enough to explain the deficit. I, as such, assume it is memory latency biting back as well.Ryzen have HT bug in games -10-15% performance drop.
Gamer'sNexus tested without smt, and it only made a real difference in one game.
http://www.gamersnexus.net/hwreview...review-premiere-blender-fps-benchmarks/page-7