For your viewing pleasure this weekend are benchmarks of TrueOS 20160831 (the rolling-release distribution formerly known as PC-BSD), DragonFlyBSD 4.6, GhostBSD 10.3, FreeBSD 11.0-RC2, and PacBSD 20160809 (formerly known as Arch BSD) all benchmarked from the same system! Plus for reference to the Linux numbers are Ubuntu 16.04.1 LTS and Clear Linux 10040 being compared to these BSDs on the same tests and hardware.
All of these tests were performed on a system with an Intel Xeon E5-2609 v4 Broadwell-EP processor and backed by an MSI X99A WORKSTATION motherboard, kind thanks to MSI Computer for sending out that motherboard and RAM. There was 16GB of DDR4 memory used during testing, OCZ TRION 150 120GB SSD, and GeForce GTX TITAN X for graphics, although the tests for this article were primarily concerned about the overall processing performance.
This is our first time comparing TrueOS, DragonFlyBSD, GhostBSD, FreeBSD, and PacBSD in one article. OpenBSD and NetBSD, among others, were attempted but those attempts were foiled. More details on the BSD testing experience with this modern Intel Xeon system can be found from the earlier article: Trying Out Eight BSDs On A Modern PC: Some Are Smooth, Others Troublesome.
Each operating system was cleanly installed on this same system and the tests then carried out while trying to stick to all of the defaults / out-of-the-box experience. With open-source operating systems being essentially tunable to an endless amount, the defaults are the best we can measure for a realistic out-of-the-box experience. If people would like stock vs. tuned OS benchmarks, as said many times before, if there is any concise Wiki guide to follow or other automated configuration scripts, I would be happy to re-test and do some additional comparison tests for those interested in the tuning potential of a particular OS.
All of the benchmarks for this article were run in a fully-automated and reproducible manner using the open-source Phoronix Test Suite benchmarking software.