Temperatures in coolant are unstable because of GPU activity.
Temperatures in coolant are unstable because of GPU activity.
Alright, I don't have any background there, so I'm going with a laugh.
This is totally new for me—time to drop it! 😂
Coolant temperature in AIOs only shows small changes, so if it feels too high, it’s likely due to warm air inside the case or poor airflow through the radiator. Fixing this could mean adding more exhaust or improving the intake/exhaust balance and flow. You’ll need to try different settings unless you share photos for better guidance. Otherwise, detailed advice will be hard to give.
The temperatures rise noticeably when you run only a CPU task, according to Cinebench23. Your CPU temperatures during Cinebench23 remain stable, but when you add a GPU benchmark (Furmark), the GPU temperatures spike significantly. The GPU heat doesn’t seem to affect the CPU coolant temperature beyond normal levels, even in a tightly sealed environment. It’s possible your sensors are malfunctioning and not reporting accurate readings.
Several additional tests confirmed my suspicion that iCUE was the issue. The pump/fans default settings are "quiet," and they don’t increase from there when iCUE isn’t active. I might file a support ticket with Corsair, as it should still display CPU temperatures without software running, but this is the only consistent pattern observed.
I'm using a specific Corsair AIO. Thanks for clarifying—I didn't catch it earlier. EDIT: It's listed on your profile. EDIT 2: @Quinten0508 Are you running the Commander CORE? And is the pump connected to it? From the Quick Start Guide, the Commander CORE depends on iCUE to manage operations, so linking the pump header to a CPU_PUMP might simplify things.