Voltage Adjustment for Graphics Processing Units
Voltage Adjustment for Graphics Processing Units
While testing GPU undervolting with Resident Evil 2 Remake, I successfully prevented system crashes by adjusting voltage levels. However, after several minutes of gameplay at maximum graphics settings—including a 200% image quality boost—the game began to crash. My query is this: would adjustments made during undervolting or overclocking indicate a need to revise those settings, or is this solely due to the game's demanding requirements exceeding my GPU’s capabilities?
(Note: This issue only occurs within the game itself, not with my computer as a whole.)
Intel i9 9900k
Gskill Trident Z 16GB
ASRock Z390 Phantom Gaming SLI/AC
AMD Radeon VII 16GB
Corsair RM 850x PSU
Intel Core i9-9900k
G.Skill Trident Z16
ASRock Z390 Phantom Gaming SLI/AC
AMD Radeon VII 16GB
Corsair RM850x power supply
These appear to be excellent parts. Could you share what temperatures you're experiencing during gameplay?
Using premium visuals,
Mboard temperature = 40 degrees Celsius
CPU temperature = remains below 51 degrees Celsius
Radeon VII: Typical reading is 43 degrees Celsius, with a maximum of 50 degrees Celsius
Radeon VII junction temperature: Average is 49 degrees Celsius, reaching a maximum of 59 degrees Celsius
At ultra graphics settings without scaling,
Mboard temperature = 42 degrees Celsius
CPU temperature = remains below 51 degrees Celsius
Radeon VII: Average reading is 51 degrees Celsius, peaking at 57 degrees Celsius
Junction temperature: Average is 60 degrees Celsius, reaching a maximum of 80 degrees Celsius
With doubled image quality and maximum graphics:
Mboard temperature = 44 degrees Celsius
CPU temperature = remains below 51 degrees Celsius
Radeon VII: Average reading is 53 degrees Celsius, peaking at 63 degrees Celsius
Junction temperature: Average is 64 degrees Celsius, reaching a maximum of 91 degrees Celsius
Regarding these configurations, I understand that they are unnecessarily elevated; I implemented this to evaluate the stability of my voltage adjustments during stress testing for my graphics card. I’m familiar with what you’re saying, thank you nonetheless.
These temperatures appear to be quite chilly. Examine your CPU and GPU activity with MSI Afterburners OSD. It's recommended to conduct load testing on your hardware utilizing Furmark or Prime95, as this will provide more detailed information than a typical game. If the system is simply failing, ensure you validate the game’s files by right-clicking on the game in Steam and selecting properties, then choosing the ‘Local Files’ tab and clicking ‘Verify integrity of game cache.’
I attempted to utilize MSI Afterburner, however it appeared to clash with my AMD Wattman’s manual fan control settings. HWMonitor and Wattman provided me with excessive information, though I retained Kombuster because it’s unnecessary to employ Afterburner for the stress test.
Nevertheless, I maintained the GPU clocks at their original values and applied a minor voltage reduction during the final test, which didn't result in a crash – however, I suspected the issue stemmed from the game itself. I overlooked file verification within Steam, and Resident Evil 2 was lacking two files, after which the game functioned correctly, even demonstrating slightly improved performance at a 200% image quality setting. I will continue to maintain a lower image quality though. Thank you for your suggestions; I haven’t yet explored Furmark or Prime95, and I plan to investigate those tools during my next testing session.