The GPU suddenly fails to manage prior timers.
The GPU suddenly fails to manage prior timers.
Hi everyone.
Recently I began using 3D Mark after installing a new CPU cooler and overclocking my CPU, only to discover that my GPU suddenly couldn't cope with the clocks it handled before. 3D Mark would crash right away during the second graphics test. I don’t understand why this is happening and even after lowering the clock significantly, crashes still occur occasionally in 3D Mark. I’ve tried reinstalling, but it doesn’t solve the issue.
Previous graphics score:
https://www.3dmark.com/3dm/44917035?
Current graphics score:
https://www.3dmark.com/3dm/45009122?
Please disregard the CPU score as it has changed a lot because of the overclock.
Help would be appreciated,
Cheers.
Fixed the problem 🙂
Adjusted core clock to +85 and memory to +450.
Link: https://www.3dmark.com/3dm/45101081? Not as good as before but sufficient for me!
CPU: AMD Ryzen 5 2600 (OC @ 4.2 GHz)
CPU Cooler: Cooler Master MasterLiquid ML240R RGB AIO
GPU: MSI GeForce RTX 2060 Gaming Z 6GB
RAM: Corsair Vengeance RGB PRO 16GB (2x8GB) DDR4 2666MHz (OC @ 3200MHz)
MOBO: MSI B450-A PRO AM4 ATX
SSD: Samsung 860 EVO 500GB 2.5" SATA III SSD
PSU: Corsair CX550M 550W 80 Plus Bronze Semi-Modular
Have you attempted to revert the CPU overclock to its earlier setting?
I lowered it back to stock and it's fine but I just can't seem to figure out why it can't handle the previous overclock that I had. I used an electric meter to see the power draw of the PC under both CPU and GPU at 100% load, and it was 420 watts, so the PSU isn't a problem.
It seems the CPU might have been limiting the performance of the overclocked GPU just enough to prevent overheating, so you only had to push it a bit further. It was definitely close to the limit, but it worked out.
Performed OCCT for power testing; GPU remained near 60°C and CPU reached about 83°C maximum. I understand the GPU temperatures are acceptable, but the CPU appears to be nearing unsafe levels, especially since OCCT is a worst-case scenario.
Fixed the problem 🙂
Adjusted core clock to +85 and memory to +450.
Link: https://www.3dmark.com/3dm/45101081? Not as good as before but sufficient for me!