results.cbenchsuite.org

cbenchsuite.org

v3.18 | cbenchsuite

http://cbenchsuite.org/tag/v3-18

Results and development of Linux benchmarking tool. Linux Kernel benchmarks v3.15 – v3.19. Here are the benchmarks for kernel versions v3.15 to v3.19. I could not see a lot of significant performance changes compared to v3.18. The results of the fork benchmark got a bit better for v3.19. Indeed they are better for more than 2 threads than all releases before. Surprisingly v3.17 fork() was faster for 1 and 2 threads. For more information and plots go to one of the result links listed above. The benchmark ...

cbenchsuite.org

v3.16 | cbenchsuite

http://cbenchsuite.org/tag/v3-16

Results and development of Linux benchmarking tool. Linux Kernel benchmarks v3.15 – v3.19. Here are the benchmarks for kernel versions v3.15 to v3.19. I could not see a lot of significant performance changes compared to v3.18. The results of the fork benchmark got a bit better for v3.19. Indeed they are better for more than 2 threads than all releases before. Surprisingly v3.17 fork() was faster for 1 and 2 threads. For more information and plots go to one of the result links listed above.

cbenchsuite.org

v3.18-ck1 | cbenchsuite

http://cbenchsuite.org/tag/v3-18-ck1

Results and development of Linux benchmarking tool. Brain Fuck Scheduler performance, Kernel version 3.18. You may have already heard of the brain fuck scheduler. The benchmark results are quite interesting. I recommend looking through all results and not just the highlights. 7zip compression benchmark with different amount of threads. Forks within a limited amount of time with a given number of threads. Time used to compile a kernel defaultconfig. Hackbench parallel message passing benchmark.

cbenchsuite.org

v4.0-rc1 | cbenchsuite

http://cbenchsuite.org/tag/v4-0-rc1

Results and development of Linux benchmarking tool. Linux Kernel Benchmarks v4.0-rc1. First v4.0 release candidate is out and here are the benchmarks in comparison to v3.19. The fork syscall got about 1% slower in 4.0-rc1 with specific number of threads forking in parallel. Hackbench without the usage of pipes got a bit faster in 4.0-rc1. At the same time is hackbench with pipe usage significantly slower than before. February 27, 2015. Proudly powered by WordPress.

cbenchsuite.org

v3.14 | cbenchsuite

http://cbenchsuite.org/tag/v3-14

Results and development of Linux benchmarking tool. Linux Kernel benchmarks v3.13-v3.18 intel i7. These plots show the results for benchmarks that were executed on an quad core i7-2xxx series. Kernel versions from v3.13 to v3.18 were used. Number of forks that could be executed within a limited amount of time. Hackbench is a benchmarking tool that sends messages between massive amounts of threads. Runtime of the pipe benchmark. February 4, 2015. Proudly powered by WordPress.

cbenchsuite.org

v3.19-rc3 | cbenchsuite

http://cbenchsuite.org/tag/v3-19-rc3

Results and development of Linux benchmarking tool. Kernel v3.18-v3.19-rc7 benchmarks. No big surprises in these benchmarks of Linux Kernel 3.18 and a couple of v3.19-rcs. Hackbench performance for 90 groups with pipe. Performance getting slightly better in v3.19-rc. Forks within a given amount of time for 16 threads. February 4, 2015. Proudly powered by WordPress.

cbenchsuite.org

v3.19-rc7 | cbenchsuite

http://cbenchsuite.org/tag/v3-19-rc7

Results and development of Linux benchmarking tool. Kernel v3.18-v3.19-rc7 benchmarks. No big surprises in these benchmarks of Linux Kernel 3.18 and a couple of v3.19-rcs. Hackbench performance for 90 groups with pipe. Performance getting slightly better in v3.19-rc. Forks within a given amount of time for 16 threads. February 4, 2015. Proudly powered by WordPress.

cbenchsuite.org

v3.17 | cbenchsuite

http://cbenchsuite.org/tag/v3-17

Results and development of Linux benchmarking tool. Linux Kernel benchmarks v3.15 – v3.19. Here are the benchmarks for kernel versions v3.15 to v3.19. I could not see a lot of significant performance changes compared to v3.18. The results of the fork benchmark got a bit better for v3.19. Indeed they are better for more than 2 threads than all releases before. Surprisingly v3.17 fork() was faster for 1 and 2 threads. For more information and plots go to one of the result links listed above.

cbenchsuite.org

Linux Kernel benchmarks v3.15 – v3.19 | cbenchsuite

http://cbenchsuite.org/2015/02/linux-kernel-benchmarks-v3-15-v3-19

Results and development of Linux benchmarking tool. Linux Kernel benchmarks v3.15 – v3.19. Here are the benchmarks for kernel versions v3.15 to v3.19. I could not see a lot of significant performance changes compared to v3.18. The results of the fork benchmark got a bit better for v3.19. Indeed they are better for more than 2 threads than all releases before. Surprisingly v3.17 fork() was faster for 1 and 2 threads. For more information and plots go to one of the result links listed above. Six +.

cbenchsuite.org

Linux Kernel Benchmarks v4.0-rc1 | cbenchsuite

http://cbenchsuite.org/2015/02/linux-kernel-benchmarks-v4-0-rc1

Results and development of Linux benchmarking tool. Linux Kernel Benchmarks v4.0-rc1. First v4.0 release candidate is out and here are the benchmarks in comparison to v3.19. The fork syscall got about 1% slower in 4.0-rc1 with specific number of threads forking in parallel. Hackbench without the usage of pipes got a bit faster in 4.0-rc1. At the same time is hackbench with pipe usage significantly slower than before. February 27, 2015. Leave a Reply Cancel reply. Your email address will not be published.