Я пробовал и понял, что в компиляции ядра половину времени занимают однопоточные и дисковые операции.
У них какой-то сферический тест видимо.
Python Linux 4.4.2 Kernel Compile Benchmark
This is one of the most requested benchmarks for STH over the past few years. The task was simple, we have a standard configuration file, the Linux 4.4.2 kernel from kernel.org, and make the standard auto-generated configuration utilizing every thread in the system. We are expressing results in terms of compiles per hour to make the results easier to read.
Here, the DDR5 transition, along with the newer core architectures, are helping Intel a lot. This is a workload where we were a bit surprised with how well the Intel parts fared. Something that is going to be a theme is that we have the dual 96-core AMD EPYC 9654 figures in our charts. This is a benchmark example of where we had to split the workload up above 64 cores into two instances because huge portions of the chip were sitting idle. That goes back to the need to run a mix of workloads on modern chips to really show what they can do.
Перевода у меня нет, поэтому в оригинале.