Core clock changing

dmFT · December 23, 2019

Hello! There is something that I've noticed recently and I just want to ask and see if there isn't anything wrong. So I just recently got a new PC with a RTX 2080 Super and Ryzen 5 3600 and I've done benchmarking with timespy on 3D Mark and I get different results and I noticed that my core clock on my GPU changed from 1950 to 1930 from one test to another.

My best result with current setup: https://www.3dmark.com/spy/9828617

My worst result with current setup: https://www.3dmark.com/spy/9486428

I just keep on getting worse results all the time and I just want to see so that nothing is wrong with it. It's not a very big deal and not a very big difference on the scores but I just think it's weird.

LIGISTX · December 23, 2019

7 minutes ago, dmFT said:

Hello! There is something that I've noticed recently and I just want to ask and see if there isn't anything wrong. So I just recently got a new PC with a RTX 2080 Super and Ryzen 5 3600 and I've done benchmarking with timespy on 3D Mark and I get different results and I noticed that my core clock on my GPU changed from 1950 to 1930 from one test to another.

My best result with current setup: https://www.3dmark.com/spy/9828617

My worst result with current setup: https://www.3dmark.com/spy/9486428

I just keep on getting worse results all the time and I just want to see so that nothing is wrong with it. It's not a very big deal and not a very big difference on the scores but I just think it's weird.

1950 to 1930 is nearly negligible. That is such a small change, you would never be able to tell in a game.

That said, to explain why its happening, its almost certainly due to the card heating up. As the card gets warmer, GPU boost dials the clocks back a bit to stay within the temp specs nvndia sets. You can override them, or increase fan speed to let it boost higher for longer etc, but this is not anything to worry about. This is 100% normal and intended results.

Also, your 3d mark scores are basically identical. Your lowest is .4% lower than your highest.... If we extrapolate that to a game and FPS in said game, if your running at 100 FPS, your would now be running at 99.6 FPS. lol.

.Apex. · December 23, 2019

14 minutes ago, dmFT said:

I just keep on getting worse results all the time and I just want to see so that nothing is wrong with it. It's not a very big deal and not a very big difference on the scores but I just think it's weird.

As @LIGISTX said, GPU Boost is related to temperature, as you repeat the test the card gets a bit hotter and GPU Boost has to adjust because of that, 1930mhz is what you would probably eventually settle on in games, try setting the fan speed manually to 100% and you would probably get an even higher score, 0.4% is nothing to be concerned about.