Understood. I prefer to get multiple, shorter samples, to show the effect and progression of throttling over time. A single 10-minute sample will hide the peak (ie, non-fully throttled performance) in its "smoothed" average.
I have published a small and simple Windows batch file for running Cinebench multiple times and aggregating the results into a single report. This can be done to measure the run-to-run performance variation for systems that thermal throttle.
Sample output:
C:\>CinebenchBatch 3
Cinebench Run #1...
I have written a script that lets you to set the Intel processor PL1/PL2 power levels under Linux. Like ThrottleStop, this script also locks the MMIO register, which is necessary to prevent newer system firmware from modifying the running PL1/PL2 values.
Note that secure boot must be disabled...
Thanks, missed that. They isolated cores using affinity masks, whereas I'm removing the cores' accessibility from the kernel. Probably doesn't make a difference in practice but I'll look into it.
I found this after @eek2121 mentioned it...
I was curious about exactly how the Intel 12th gen efficiency cores compare to the performance cores, so I wrote a script that runs Geekbench 5 against each individual core. I chose Geekbench for this test because it offers a nice mix of real-world algorithms rather than an arbitrary synthetic...
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.