3 ms·
Speedometer is not really a "Javascript" benchmark, it's a browser benchmark, all the actual "heavy" lifting is DOM operations. It's absolutely useless as a be
by zeroname 8y ago
Speedometer is not really a "Javascript" benchmark, it's a browser benchmark, all the actual "heavy" lifting is DOM operations.
It's absolutely useless as a benchmark for hardware, because it's testing wildly different codebases. Safari on iOS is not the same as Safari on Mac OS. DOM performance on iOS Safari has always been remarkable compared to Desktop/Android browsers.
It's pretty sad to see an entire thread of clueless people (including DHH) talking about how amazing the hardware advances are when it's all software.
- MiddleEndian 8y agoIt is pretty damning of the software though.
- ShinTakuya 8y agoI mean mobile software has had a lot more motivation to be efficient due to battery usage. Additionally it's a lot newer than most desktop browsers so it can learn from older mistakes. Firefox's recent work has effectively shown what rewriting browser components can do for performance.
- zamadatix 8y agoThe latest Firefox still scores worse than Chrome on this benchmark (on my PC at least). As much as I'd like to believe it's just really great code I'm inclined to believe it's not doing the benchmark in the same way. We're talking a mobile phone pulling a few watts getting a significantly higher score than my 5 GHz i7 with a 1080 accelerating the drawing regardless of which browser I run. Something is whacky and it's not just "it's a lot newer than most desktop browsers". Could be some relaxed timing from what the spec requires, display rules not being followed exactly, prewarming the engine with the benchmark, or some other deviation.