Linux's P-State Performance Governor Shows Unexpectedly Big Boosts For The Intel Core i9-11900K

[phoronix]

Phoronix: Linux's P-State Performance Governor Shows Unexpectedly Big Boosts For The Intel Core i9-11900K

The P-State and CPUFreq "performance" governors on Linux with desktop Intel/AMD processors can be of help for gaming and other select workloads by tending to keep the CPU clock frequencies higher than the default ondemand (CPUFreq) or powersave (P-State) governors used by nearly all Linux distributions. But with Intel's new Core i9 11900K "Rocket Lake" is a dramatic difference in power and performance between the Intel P-State performance and powersave governors than what we have seen over the years with prior generations of Intel Core processors.

https://www.phoronix.com/vr.php?view=30146

[bug77]

That's just how Intel CPUs work these days. Thermals allowing, they can tap into the last drop of juice the CPU has, at the expense of power draw. And they do reach that zone where the next 1% extra juice requires a hell of lot more power.

[zcansi]

Maybe I'm missing something but the title seems to really bury the lede:

This testing was also the first time where using the P-State performance governor induced stability problems with the wild power usage... When running the P-State powersave governor all the workloads were stable as expected. With the P-State performance governor is the first time I have found the i9-11900K to randomly reboot under Linux. LuxCoreRender (CPU-based rendering mode) and SVT-HEVC (Intel's optimized H.265 CPU video encoder) were two workloads that were causing the system to suddenly reboot when thrashing the processor. This didn't occur with the powersave governor but happened reliably with the performance governor multiple times and had to blacklist those test profiles from the run in order to be able to successfully complete all of the benchmarks without any reboots under load.

Maybe the title should be "Linux P-State Performance Governor Overclocks CPUs Beyond TDP, Causing Repeatable Crashes"?

[ddriver]

LOL, this thing doesn't stay within its rated power envelope even on powersave....

Hey, maybe intel is secretly setting the tdp based on the power use of a SINGLE cpu core load?

[numacross]

Intel defines TDP at base clocks so this CPU is 125W only at 3.5GHz.

Thermal Design Power (TDP) represents the average power, in watts, the processor dissipates when operating at Base Frequency with all cores active under an Intel-defined, high-complexity workload. Refer to Datasheet for thermal solution requirements.

[ddriver]

It is the year 2021 and people are so intellectually advanced that they need to have sarcasm explicitly labeled for them.

[pracedru]

Considering the extra power used, it doesn't really make sense to enable performance mode.
Unless you really need those extra muscle.

[aufkrawall]

intel_pstate powersave is such a joke. Dear maintainers, please just force-pull the plug and let it die.

[bobbie424242]

I cannot believe Michael changed the distro defaults !

[Jumbotron]

LOL....not surprised at all by this. Charlie Demerjian over at Semi accurate.com had been warning after Comet Lake came out that Rocket Lake was going to be bad. After it did he stated that Rocket Lake should have never been released and called it "a ghastly pig of a power hog" and that it was absolutely stupid to backport 10nm CPU and GPU cores to Rocket Lake 14nm cores.

https://semiaccurate.com/2021/03/16/intels-rocket-lake-cpu-shouldnt-have-been-made/