This makes sense to me: hard to beat the HW rasterizer for blending. But what’s strange is that I’ve seen numbers that show compute being 40x faster than what I’m seeing on comparable hardware, and I can’t reproduce them…
I’m basically doing what piet-metal is doing. Major differences are (1) tiling is on CPU not GPU (but I’m only talking about rendering time here, not tiling); (2) strokes are converted to fills first on GPU instead of being handled natively.
-
-
Piet is not a branch away from PF3, so I am still not able to connect the dots of what you are saying. Having a technical discussions on Twitter is hard. I guess I'll just look at the code when the compute branch is ready to make it into master
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.