Two Years With Linux BFS, The Brain Fuck Scheduler

While in some applications the Brain Fuck Scheduler had not resulted in the workload being carried out any faster, what the tests don't measure is the responsiveness / latency of the system as seen by the user, which is one of the benefits of a good task scheduler.

With the 7-Zip compression test there was no longer a performance difference in the benchmark result, but the CPU usage was different.

The quad-core Intel Core i5 2500K CPU usage was nearly 10% lower when running 7-Zip with the Brain Fuck Scheduler over the Completely Fair Scheduler.

Above is a look at the CPU usage spanning several different tests. The average CPU usage with the Brain Fuck Scheduler was 7% lower than the Completely Fair Scheduler.

Those interested in more benchmarks can visit OpenBenchmarking.org.