Does Using GNOME On Wayland Save Power?
One of the commonly asked questions is whether using Wayland will be more power efficient or save power compared to running the same software under an X.Org Server environment. Here's a simple test of GNOME on Wayland in Fedora 21 while monitoring a laptop's battery use.
After yesterday's Fedora 21 Gaming Benchmarks: X.Org vs. XWayland To End 2014 I ran a simple comparison just looking at the system power consumption while engaging with the GNOME 3.14.2 desktop environment under X.Org Server 1.16 and when running under Wayland.
The same ASUS Zenbook with Core i7 3517U Ivy Bridge processor with HD Graphics 4000 was used for testing with 4GB of RAM and dual 128GB SanDisk SSDs. Fedora 21 was running with the Linux 3.17 kernel and Mesa 10.4.
The Zenbook ultrabook's battery usage was monitored by setting the MONITOR=sys.power environment variable prior to testing with the open-source Phoronix Test Suite benchmarking software. Monitored was the GNOME desktop when running a brief LAME MP3 encoding task to load up the CPU for some reference data points followed by the system idling and then launching a new GNOME Terminal window, launching a Nautilus window, and launching a gedit text editor window. For one minute the windows were moved around / rather idle desktop work.
Here's the result of this brief comparison:
Under Fedora 21 between GNOME on X and GNOME on Wayland, the power use really wasn't different. Ignoring the one, odd spike shown during the tests, the power consumption was basically the same between the two environments.
The CPU usage for this Core i7 Ivy Bridge ultrabook/laptop was also indifferent sans the one spike. The results were uploaded to this OpenBenchmarking.org result file.
Stay tuned for more Mir and Wayland benchmarks on Phoronix in 2015.
After yesterday's Fedora 21 Gaming Benchmarks: X.Org vs. XWayland To End 2014 I ran a simple comparison just looking at the system power consumption while engaging with the GNOME 3.14.2 desktop environment under X.Org Server 1.16 and when running under Wayland.
The same ASUS Zenbook with Core i7 3517U Ivy Bridge processor with HD Graphics 4000 was used for testing with 4GB of RAM and dual 128GB SanDisk SSDs. Fedora 21 was running with the Linux 3.17 kernel and Mesa 10.4.
The Zenbook ultrabook's battery usage was monitored by setting the MONITOR=sys.power environment variable prior to testing with the open-source Phoronix Test Suite benchmarking software. Monitored was the GNOME desktop when running a brief LAME MP3 encoding task to load up the CPU for some reference data points followed by the system idling and then launching a new GNOME Terminal window, launching a Nautilus window, and launching a gedit text editor window. For one minute the windows were moved around / rather idle desktop work.
Here's the result of this brief comparison:
Under Fedora 21 between GNOME on X and GNOME on Wayland, the power use really wasn't different. Ignoring the one, odd spike shown during the tests, the power consumption was basically the same between the two environments.
The CPU usage for this Core i7 Ivy Bridge ultrabook/laptop was also indifferent sans the one spike. The results were uploaded to this OpenBenchmarking.org result file.
Stay tuned for more Mir and Wayland benchmarks on Phoronix in 2015.
93 Comments