Battery drain test and browser running side by side

[samsam999]

Has anyone read about or tested different browser running side by side on similar spec macbooks to see which browser is most battery efficient? Or checked watts draw?

For test to be valid it would have to be ran side by side visit same website and scrolling similar.
Preferably a non-video test, a mixed test and a video-only test with multi tabs.

I saw some older tests showing that watching no Youtube the browser were drawing almost same watts but as soon as video was played Safari draw far less, then Firefox and last Chrome with about 2.5x battery draw than Safari.

 

[Krevnik]

Has anyone read about or tested different browser running side by side on similar spec macbooks to see which browser is most battery efficient? Or checked watts draw?

Some of these measurements are going to be very specific to a scenario or site as well, on top of what quirks that site might use with that browser. So just fair warning that there's going to be caveats around these sort of measurements that don't look at a number of different scenarios.

I saw some older tests showing that watching no Youtube the browser were drawing almost same watts but as soon as video was played Safari draw far less, then Firefox and last Chrome with about 2.5x battery draw than Safari.

This is one specific scenario. Odds are very good that this result was because YouTube was using H.264 with Safari and VP9 with Chrome/Firefox. VP9 is preferable to Google for licensing reasons, but at the time of the test, had no hardware decoder accessible on macOS. So the draw was really HW decoded H.264 vs SW decoded VP9.

 

[yurkennis]

Has anyone read about or tested different browser running side by side on similar spec macbooks to see which browser is most battery efficient? Or checked watts draw?

For test to be valid it would have to be ran side by side visit same website and scrolling similar.
Preferably a non-video test, a mixed test and a video-only test with multi tabs.

I saw some older tests showing that watching no Youtube the browser were drawing almost same watts but as soon as video was played Safari draw far less, then Firefox and last Chrome with about 2.5x battery draw than Safari.

I performed somewhat similar benchmarks:
1) for Chrome only, I measured memory usage
2) for Safari only, I compared power usage for Youtube video playback between M1 Pro and M1 Max

From this experience, the most tricky parts are:

• to agree which sites to consider
• what exactly comprises "scrolling similar"
• how to automate those "scrolling" (Selenium?)
• for video tests, how to maintain continuous playback and equal video resolution setting: both of them resets every time your network connection drops (which turned out to occur once or twice per day for my Internet provider). Upon connection reset, video looping becomes disabled; resolution resets to Auto