Phoronix Test Suite 2.0.0 running on GNU/Linux |
|
Original author(s) | Phoronix Media |
Developer(s) | Michael Larabel, Matthew Tippett |
Initial release | April 2008 |
Stable release | 3.6.1[1] / December 23, 2011 |
Written in | PHP |
Operating system | Linux, FreeBSD, OpenBSD, NetBSD, OpenSolaris, Mac OS X, Windows |
Size | 475 kB (base program) |
Available in | English |
Type | Benchmark |
License | GNU General Public License v3 |
Website | phoronix-test-suite.com |
Phoronix Test Suite (PTS) is a free, open-source benchmark software for Linux and other operating systems developed by Phoronix Media with cooperation from an undisclosed number of hardware and software vendors. The Phoronix Test Suite has been endorsed by sites such as Linux.com[2], LinuxPlanet[3] and has been called "the best benchmarking platform" by Softpedia[4]. The Phoronix Test Suite is also used by Tom's Hardware[5], ASELabs[6] and other review sites.
Contents |
Phoromatic is an web-based remote test management system for the Phoronix Test Suite. It does automatic scheduling of tests. It is aimed at the enterprise. It can manage multiple test nodes simultaneously within a test farm or distributed environment.
Phoromatic Tracker is an extension of Phoromatic that provides a public interface into test farms.[9] Currently their reference implementations autonomously monitor the performance of the Linux kernel on a daily basis[10], Fedora Rawhide[11] and Ubuntu[12].
Company / developer | Phoronix Media |
---|---|
OS family | Unix-like |
Working state | Current |
Initial release | August 4, 2009 |
Latest stable release | 2010.1 "Anzhofen" / February, 2010 |
Supported platforms | x86-64 |
Kernel type | Monolithic (Linux) |
Userland | GNU |
Default user interface | GNOME |
Official website | pts-desktop-live.com |
PTS Desktop Live is a stripped down x86-64 Linux distribution. It is designed for testing/benchmarking computers from a LiveDVD / LiveUSB environment.[13]
Original author(s) | Phoronix Media |
---|---|
Written in | PHP |
Type | Library |
License | GNU General Public License v3 |
Website | Phorogit |
Phodevi (Phoronix Device Interface) is a library that provides a clean, stable, platform-independent API for accessing software and hardware information.
Phoronix Certification & Qualification Suite (PCQS) is a reference specification for the Phoronix Test Suite.
Version | Codename | Date |
---|---|---|
3.8 | Bygland | TBA Q1 2012 |
3.6 | Arendal | December 13, 2011 |
3.4 | Lillesand | September 8, 2011 |
3.2 | Grimstad | June 15, 2011 |
3.0 | Iveland | February 26, 2011 |
2.8 | Torsken | August 31, 2010 |
2.6 | Lyngen | May 24, 2010 |
2.4 | Lenvik | February 2, 2010 |
2.2 | Bardu | November 16, 2009 |
2.0 | Sandtorg | August 4, 2009 |
1.8 | Selbu | April 6, 2009 |
1.6 | Tydal | January 20, 2009 |
1.4 | Orkdal | November 3, 2008 |
1.2 | Malvik | September 3, 2008 |
1.0 | Trondheim | June 5, 2008 |
On 2008-06-05 Phoronix Test Suite 1.0 was released under the codename Trondheim.[14] This 1.0 release was made up of 57 test profiles and 23 test suites.[15]
On 2008-09-03, Phoronix Test Suite 1.2 was released with support for the OpenSolaris operating system[16] and a module framework accompanied by tests focusing upon new areas[17] and many new test profiles.
Phoronix Test Suite 1.8 includes a graphical user interface (GUI) using GTK+ written using the PHP-GTK bindings.
3.4 includes MATISK benchmarking module and initial support for the GNU Hurd.
Phoronix uses Phoronix Test Suite to compare performance of different operating systems - Linux distributions, OpenSolaris and FreeBSD. Most controversies focus on the third party tests within Phoronix Test Suites and not the tests themselves. Selected controversies are covered below.
The compiler used in building the source based tests have a large impact on the results. Depending on the readers point of view this is seen as a detriment to the value of the results. However other readers may see that as a realistic expectation of performance for that particular system with that operating system.
Also, some of the tests that claim to e.g. measure file system performance, in fact measure speed of operations that never happen in real applications. An example of this can be "Threaded I/O Tester", which doesn't actually measure any I/O performance - it measures time required to read from a single empty, sparse file, which is an obviously pointless operation. It would be trivial to change this benchmark to measure something sensible - for example, reads from a file containing data - but such tests were not published by Phoronix, for reasons unknown.