Citation on the 15-30% claim? We see more like 50%, including compilation time. Sometimes much faster on certain benchmarks (Bullet IIRC?) now that float32 support has landed.
This is from a link in the article for more information. These Monster Madness fellows found Firefox gave them 33% of native performance, and Chrome 20%. Presumably IE would give you fractions of these figures.
I figured that, rather than citing figures from benchmarks, I'd go for the real world scenario that this release has given us.
Comments
Citation on the 15-30% claim? We see more like 50%, including compilation time. Sometimes much faster on certain benchmarks (Bullet IIRC?) now that float32 support has landed.
https://hacks.mozilla.org/2013/12/monster-madness-creating-g...
This is from a link in the article for more information. These Monster Madness fellows found Firefox gave them 33% of native performance, and Chrome 20%. Presumably IE would give you fractions of these figures.
I figured that, rather than citing figures from benchmarks, I'd go for the real world scenario that this release has given us.
Those figures seem to be in the context of a week long initial port. The team lead uses a figure of 50% in the video.
Also, much of these games is 3D rendering, and WebGL runs at essentially full native performance.