Stuttering in games - Frametime spikes

[Pikatchu]

I have recently moved my setup from an openbench to inside a case. The new case is Lian Li O11D mini, not the airflow one.

 

After the transition, I notice my game started to having stutters, random frame pauses. Checking with the RiverTune software, it is confirmed that my frametiming is spiking. This issue did not exist, or at least not noticeable, when it is on the openbench. The game in question is F1 23, as it is the one I had time playing since the transition. I will update the post when I had time to test other games.

 

What may have caused this issue to rise up now?

 

Some background to the setups:

1) I have been running in an openbench setup for almost an year now. The decision to move it back into a case was because the machine needs to be on 24/7. Leaving it on an openbench will require cleaning more frequently, which for most of the time I am not home near the machine.

 

2) My specs are: NH-D15 cooling a Ryzen 9 7900X with offset bracket, RTX 3080 FTW3, 64GB DDR5 6000Mhz RAM, 4TB + 1TB NVME gen 4 SSD. Case configuration for the fans are 3 bottom intake, 2 side intake, 3 top exhaust. Fan profile for the fans are 2 side intake stay at a constant speed, only top and bottom row change base on temperature of GPU.

 

3) Temperatures during gaming

Openbench - CPU: mid 70s Celsius; GPU: mid 60s Celsius; GPU Mem: low 80s Celsius (+15-20Celsius of GPU Core); Storage: 50 Celsius

In case        -  CPU: low 80s Celsius; GPU: low 70s Celsius; GPU Mem: low 90s Celsius (+15-20Celsius of GPU Core); Storage: 54 Celsius

 

My thoughts are the increase in temperature may have caused the increase in stuttering, but they are all within acceptable ranges.

[AAVVIronAlex]

Your temps look fine, the CPU is not overheating, and if it does those parts are designed from the ground-up to overclock themselves until they reach the 95C they stay at. The GPU temps look fine (core), I have never used a 3080 and did not really pay attention to memory temps on the GPUs, but that seems to be fine. They should be able to go up to 100C without too much degradation.

 

I would look forward to reseting the BIOS and then adding in the settings once more if nothing changes.

[Ebony Falcon]

Go bios and do optimised defaults 

boot back in and check again

if that fixes it ur ram xmp isn’t stable 

[Pikatchu]

9 hours ago, Ebony Falcon said:

Go bios and do optimised defaults 

boot back in and check again

if that fixes it ur ram xmp isn’t stable 

 

11 hours ago, AAVVIronAlex said:

Your temps look fine, the CPU is not overheating, and if it does those parts are designed from the ground-up to overclock themselves until they reach the 95C they stay at. The GPU temps look fine (core), I have never used a 3080 and did not really pay attention to memory temps on the GPUs, but that seems to be fine. They should be able to go up to 100C without too much degradation.

 

I would look forward to reseting the BIOS and then adding in the settings once more if nothing changes.

 

Thanks for the advices.

 

I have reset the BIOS to default and then gradually revert my settings back seems to have fixed the issue.

It is weird, as it is back to the same settings as previously and the issue no longer exists.

[Salted Spinach]

Could look into undervolting your CPU. I gave my 5800X3D an -60mV offset and it shaved almost 20C in games. It now sustains boost frequency more often than before, thanks to the thermal headroom.