Hello,

I’ve gone by way of with the answer posted by McCalpin John: https://neighborhood.intel.com/t5/Software program-Archive/Intel-Xeon-Phi-Studying-MSR/m-p/1170935

I’ve used IA32_THERM_STATUS (MSR 0x19c) to measure instantaneous temperature for a protracted length by operating an actual world benchmark. I’ve hooked up a picture of temperature vs execution timeline. I’ve seen steep adjustments in temperature worth inside just a few seconds which isn’t alleged to occur as temperature needs to be rise and fall very steadily I imagine. Can anybody please verify whether it is right or not by wanting an hooked up picture.

Thanks!