benchmark and priority

Pkirk618

Supreme [H]ardness
Joined
Sep 12, 2002
Messages
5,708
I haven't upgraded in years (until this week) so idk if this is true. Is manipulating a program's (benchmark program) priority considered cheating?

I noted I get a considerable boost when running at a higher priority.
 
Last edited:

TheSlySyl

Gawd
Joined
May 30, 2018
Messages
772
In my super unscientific and anecdotal experience. It shouldn't make a difference unless you're, for some reason, benchmarking with a bunch of background tasks running.
 

LukeTbk

Limp Gawd
Joined
Sep 10, 2020
Messages
453
That sound similar to considering cheating to run the benchmark on a fresh Windows installation or with an abnormally low amount of background task/service/antivirus running versus a regular user.

If the goal of a benchmark is to compare how you machine is doing versus expectation of what it should be able to do for the hardware, and that you comparison is on freshly installed windows machine with nothing else but the benchmark because that what reviewer usually do, that seem the things to do.
 

Pkirk618

Supreme [H]ardness
Joined
Sep 12, 2002
Messages
5,708
backgrounds absolutely reduce the score -- this I can confirm. However, I can't help but think those who take these benchmarks real serious don't eliminate choke points big or small. When I bench in with anything higher than normal priority my scores just a quite a bit, especially with the highest priority. Sometimes when I submit a score and I see similar spec'd systems score higher I wonder what's wrong with my PC.
 

primetime

Supreme [H]ardness
Joined
Aug 17, 2005
Messages
6,647
I wouldnt call it cheating but it is what it is (compared to not upping priority)
 

Jonnycat99

Limp Gawd
Joined
Nov 9, 2006
Messages
155
Hardware benchmarking is testing hardware, so it makes sense to take as much of the OS out of the way as possible when performing the test.
 

TheSlySyl

Gawd
Joined
May 30, 2018
Messages
772
backgrounds absolutely reduce the score -- this I can confirm. However, I can't help but think those who take these benchmarks real serious don't eliminate choke points big or small. When I bench in with anything higher than normal priority my scores just a quite a bit, especially with the highest priority. Sometimes when I submit a score and I see similar spec'd systems score higher I wonder what's wrong with my PC.
Oh I know background tasks reduce the score, but I also shut them off if I'm going for a real benchmark.
From my experience, the difference between shutting off my Minecraft Server, Plex Server, Rainmeter skins and the Afterburner overlay is about 200 pts in Time Spy and about 400 pts in Cinebench R20.

Rainmeter is by far the biggest resource hog of those, probably because I have it it updating so frequently. While gaming the difference is negligeable. (1-2%, which is equivalent to the 200 points that I lose in Time Spy.)
 

Pkirk618

Supreme [H]ardness
Joined
Sep 12, 2002
Messages
5,708
What bench? I've not seen anything close to a "considerable" difference in my years of benching.
r20 mostly. Couple of hundred points if I recall. This, and I thought something was wrong with my system so that prompted me to ask the question. Since then, I just disable some programs running the background and I'm content with my scores. Also I think CPU-Zs benchmark better if nothing is competing for clocks.
 

primetime

Supreme [H]ardness
Joined
Aug 17, 2005
Messages
6,647
What's a nice boost??

I just tried it and it made a difference of 36 pts in a 7000+ pt score. That amounts to a whopping .5%. Omg!
idk maybe different pc's react differently? mine does thing just from changing priority
Capture.JPG
 
Top