Arch Linux vs. Antergos vs. Clear Linux vs. Ubuntu Benchmarks

Written by Michael Larabel in Operating Systems on 17 January 2018 at 07:40 PM EST. Page 1 of 4. 20 Comments.

Last week when sharing the results of tweaking Ubuntu 17.10 to try to make it run as fast as Clear Linux, it didn't take long for Phoronix readers to share their opinions on Arch Linux and the request for some optimized Arch Linux benchmarks against Clear Linux. Here are some results of that testing so far in carrying out a clean Arch Linux build with some basic optimizations compared to using Antergos Minimal out-of-the-box, Ubuntu Server, and Clear Linux.

Tests this time around were done on the Intel Core i9 7980XE system with ASUS PRIME X299-A motherboard, 4 x 4GB DDR4-3200 Corsair memory, GeForce GTX 750, and Corsair Force MP500 120GB NVMe solid-state drive. The system with 18 cores / 36 threads does make for quick and easy compiling of many Linux packages.

The operating systems tested were:

- Ubuntu 17.10 Server as a reference point with all available system updates, which does take it to the Linux 4.13.0-25-generic kernel that has Kernel Page Table Isolation (KPTI) for fending off Meltdown. The Ubuntu 17.10 Server was used as a basic up-to-date non-rolling-release Linux distribution reference point given the popularity of Ubuntu.

- Antergos Minimal 18.1 with all available updates, taking it to the Linux 4.14.13-1 kernel and GCC 7.2.1. The 4.14.13 point release is post KPTI merging so that support is in place but like Ubuntu and Arch Linux does not have any Retpoline support for Spectre. Antergos Minimal was tested out-of-the-box for getting a clean, "normal" Arch-based experience for those wishing to see have an easy-to-use desktop distribution that is quick and easy to reproducibly deploy (why we prefer Antergos when normally doing Arch-based testing, due to it at least being easily reproducible OS experience and providing a sane default experience).

- A clean installation of Arch Linux x86_64 while taking into account basic optimizations ensuring CFLAGS/CXXFLAGS as well as the compiler flags in makepkg.conf were "-O3 -march=native" before using makepkg, etc. Over the default experience was also switching from the powersave P-State governor to performance. Arch Linux is up to the Linux 4.14.13 kernel (including KPTI) and GCC 7.2.1 as the key components worth mentioning for this round of testing, more details in the system table below.

- Clear Linux out-of-the-box in its latest rolling-release form of Clear 20350 and right now it also happens to be at Linux 4.14.13 with GCC 7.2.1. Though worth pointing out in Clear's case is they have already back-ported the Retpoline patches to its kernel as well as the Retpoline GCC support, so this is the only distribution tested in this comparison both with full KPTI support and full generic retpoline protection.

Arch Linux vs. Antergos vs. Clear Linux vs. Ubuntu

Via the Phoronix Test Suite a range of mostly system/CPU tests were carried out for seeing how well the manual Arch Linux install compares to Antergos Minimal out-of-the-box, how that compares to Ubuntu, and if the basic Arch Linux changes are enough for it to run closer to Intel's performance-oriented Clear Linux.


Related Articles