• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

About that HPET setting...

zer

zer

zer

Associate
Joined
6 Jan 2004
Posts
461
Location
Greece
Hi there!
I am totally confused about all this noise concerning Spectre,Meltdown patches and lately, HPET at default or forced on setting...English not being my mother toungue doesnt help either...
So, guys: What is the punchline here?
Someone with Intel CPU (i have i5 8400) and Win 10 should check some settings, besides having the latest win 10 updates installed (which is all i do)? PC used mainly for gaming, but web browsing, power point creation etc as well ...
Many thanks!
 
Last edited:
humbug said:
So it only makes about a 3% difference.
Click to expand...

Games seem to be different.

69511297df1e37a4877587e65c233b849f1c475061f81d1cc7fb6ff7c1e5f6a81cda9966.jpg
 
Thats still only 3% with Tomb Raider only 7%, can you show me 5 games where the same is true or is that just a one off?
 
humbug said:
Thats still only 3% with Tomb Raider only 7%, can you show me 5 games where the same is true or is that just a one off?
Click to expand...

Just quick testing, Far Cry 5 is defo a change I have benched that so much lately so I know any change in that is correct.

Am not here looking to jump to the hills on performance :D a gain is a gain end of the day. Am more interested in seeing how games play. People are reporting smooth performance from better CPU load to the GPU. Squad will be a nice test for this but sadly no in game benchmark for that.
 
shankly1985 said:
Just quick testing, Far Cry 5 is defo a change I have benched that so much lately so I know any change in that is correct.

Am not here looking to jump to the hills on performance :D a gain is a gain end of the day. Am more interested in seeing how games play. People are reporting smooth performance from better CPU load to the GPU. Squad will be a nice test for this but sadly no in game benchmark for that.
Click to expand...

The problem is some people like to proport that its 30% +, its nothing like that is it?
 
shankly1985 said:
Some of the big hitters from that link above I sadly dont have installed. I do have GTA 5 though i can test.
Click to expand...

Anand's revised HPET runs they claim they get 76% higher performance for RoTR on the 8700K.

I just have a hard time reconciling this, i ran RoRT on my overclocked GTX 1070 and got around 100 FPS, this with the GPU mostly maxed out.

That's about the same as Anand reported on their initial run with the GTX 1080, now after getting complaints that Ryzen 2 didn't look far enough behind Coffeelake they say the problem was HPET and they are getting 76% higher performance with it off, i know the GTX 1080 is already maxed out at around their original result, because i have a similar GPU and the game, its not possible to get 76% more performance out of it, 10% at best.

Now if they had said 10%, fine.... but they have gone to ridiculous extremes. its clearly nonsense.
 
Last edited:
Everyone is really overthinking this. Windows 10 does not use HPET unless you tell it to. If you haven't done that, it'll be off and you're fine. Just run the command to check in Start -> cmd:

bcdedit /deletevalue useplatformclock
Click to expand...
If it errors, HPET is already disabled.
 
DragonQ said:
Everyone is really overthinking this. Windows 10 does not use HPET unless you tell it to. If you haven't done that, it'll be off and you're fine. Just run the command to check in Start -> cmd:


If it errors, HPET is already disabled.
Click to expand...

I did this and got error, but this doesn't explain why I seeing gains? I disabled HPET in device manager.
 
DragonQ said:
You are messing with timers and expecting accurate results from benchmarking software that relies on timers?
Click to expand...

This, and its actually turned on on Ryzen CPU's because those times were causing them to report much higher clocks than they actually were.

It also explains Anad's clearly wrong revised performance figures.
 
DragonQ said:
You are messing with timers and expecting accurate results from benchmarking software that relies on timers?
Click to expand...

No if HPET is disabled on Windows 10 by default then I should see zero change in my results. Thats what i dont understand. Is it possible that Windows 10 applications can use HPET on the fly while the OS doesn't?
 
shankly1985 said:
No if HPET is disabled on Windows 10 by default then I should see zero change in my results. Thats what i dont understand. Is it possible that Windows 10 applications can use HPET on the fly while the OS doesn't?
Click to expand...
That's exactly the point. Applications that want to use HEPT can do so, but you are forcing it disabled so they can't. If they are expecting to be able to use HPET but can't, then you can't trust the benchmark results they give.
 
You must log in or register to reply here.
Back
Top Bottom