Discussion Is this something I should really worry about?

[drea.drechsler]

https://www.tomshardware.com/reviews/amd-ryzen-clock-bug-benchmark-scores,6312.html

It just seems to me only a problem to the rarefied world of comptetitive over-clockers and bench-markers running on LN2 where any advantage will be exploited for even a tiny "improvement" in numbers.

Will real-world results vary so greatly that it exceeds margin-of-error? And then, if I run some kind of utility that messes with the system clock to make results appear WAY good, who am I lying to except myself?

 

[DMAN999]

Not unless you are part of an Extreme Overclocking Team and need to have your Extreme OC's verified.

 

[drea.drechsler]

Not unless you are part of an Extreme Overclocking Team and need to have your Extreme OC's verified.

That's kinda what I'm thinking. But I'm also curious about this because you can't really look at clocks to know if your Ry3k is performing as it should. You really have to run a benchmark (CB20 being pretty popular) to compare to what various review sites and others get. So long as I'm within margin of error, I'm happy it's OK even though I so rarely see 4.4Ghz boosting.

I also noted that my BCDEDIT store did not have HPET enabled...so I enabled it. It didn't make any difference with CB20 scores same as previous. But I suspect that's not good enough for competition where one point can make the margin of victory.

 

[Ninjawithagun]

Enabling or disabling HPET usually won't help with CPU timing issues, and in fact can cause more harm than good by changing the default. The only realy reason to enable HPET is if you are running specific benchmarks that otherwise don't correctly utilize the CPUs onboard timing protocol. For Ryzen, it's a definite "no no" to mess with HPET:

https://www.pcworld.idg.com.au/arti...hpet-can-make-your-ryzen-computer-run-slower/

 

[drea.drechsler]

Enabling or disabling HPET usually won't help with CPU timing issues, and in fact can cause more harm than good by changing the default. The only realy reason to enable HPET is if you are running specific benchmarks that otherwise don't correctly utilize the CPUs onboard timing protocol. For Ryzen, it's a definite "no no" to mess with HPET:

https://www.pcworld.idg.com.au/arti...hpet-can-make-your-ryzen-computer-run-slower/

mheh. I'm very aware there's been a raging debate about HPET since Ryzen 1000 release... and really since well before that. I'm not sure I've ever noticed any difference when I enable or disable it whatever processor I was running at the time (that includes a P4, Core2Duo, 6300, 1700 and now my 3700X).

I really only did it this time in response to that article, as a test. It was an easy toggle to enable, and equally easy to disable. If anything pops up I'll go back and disable it.

 

[Ninjawithagun]

mheh. I'm very aware there's been a raging debate about HPET since Ryzen 1000 release... and really since well before that. I'm not sure I've ever noticed any difference when I enable or disable it whatever processor I was running at the time (that includes a P4, Core2Duo, 6300, 1700 and now my 3700X).

I really only did it this time in response to that article, as a test. It was an easy toggle to enable, and equally easy to disable. If anything pops up I'll go back and disable it.

The most problematic issue that I noticed in several recent HPET on/off reviews was the stuttering during gameplay. Some games didn't exhibit the issue, while other games were almost crippled by the high amount of stuttering and frame skipping. Whether or not you 'think' it's an issue or not is irrelevant. It's been proven to be a problem and not an end all solution.

 

[drea.drechsler]

... It's been proven to be a problem and not an end all solution.

Probably what most escapes me is the problem enabling...or disabling...purports to solve.

I'll stay alert to stuttering though.