Linux Kernel Benchmarks Of 2.6.24 Through 2.6.39

With the recent look at the major Linux power regressions taking place within the Linux kernel, some initially wondered if the increase in power consumption was correlated to an increase in system performance. Unfortunately, it is clear now that is not the case. With that said though, here's some performance benchmarks of all major kernel releases going back to Linux 2.6.24 and ending with the Linux 2.6.39 kernel.

This is not our first time doing such a large kernel comparison, but recently there were the five years of Linux kernel benchmarks. This though was with an older user-space and in a virtualized environment to be able to handle the kernel support going back years with most of the hardware within our labs not being supported that far back. In this article, the testing is being done on actual hardware. The hardware under test is an Intel Core 2 Duo E8400 overclocked to 3.87GHz, a Gigabyte EP45T-DS3R motherboard, 2GB of DDR3 system memory, 160GB Western Digital SATA HDD, and ATI Radeon HD 4850 graphics.

Originally, this comparison was slated to be much larger, but after it becoming clear that system performance increases (or decreases) was not related to the increased power consumption in the recent kernels, the test load was reduced. Additionally, I had to leave this testing be when heading back over to Europe, so it was scaled back to just one system and a variety of tests.

These tests though are being published today for reference. Those interested can also further extend the testing themselves by using the Phoronix Test Suite and OpenBenchmarking.org. The result file containing these results is 1104261-GR-LINUXKERN05.

For easy reproducibility, each major kernel was obtained from the Ubuntu mainline kernel PPA in order to use vanilla releases without any extra distribution-supplied patches, etc. The OS setup was Ubuntu 8.04.4 LTS x86_64 with GNOME 2.22.3, xf86-video-vesa driver, X.Org Server 1.4.0.90, Mesa 7.0.3-rc2, GCC 4.2.4, and an EXT3 file-system.