Open Bug 985262 Opened 12 years ago Updated 2 years ago

Slower performance than Chrome on Voronoi polygon computations

Categories

(Core :: JavaScript Engine, defect, P3)

x86
macOS
defect

Tracking

()

People

(Reporter: bzbarsky, Unassigned)

References

(Blocks 1 open bug)

Details

Steps to reproduce: 1) Load http://www.raymondhill.net/voronoi/rhill-voronoi-demo1.html 2) Check the number in the textfield from "100" to "10000" 3) Click the "Benchmark" button. I see us end up with about execution 130ms, while Chrome is at about 100ms. This is on Mac. The claim is that the time measured does NOT include the canvas drawing bits, so ought to be pure JS time.
On Windows 8 Nightly its a bit close to call: Nightly: average exec time for 10000 sites = 103.3 ms (10.3 µs/site), fastest = 92 ms, slowest = 264 ms. Chrome: average exec time for 10000 sites = 97.5 ms (9.8 µs/site), fastest = 71 ms, slowest = 139 ms. It's not really related to this bug but I noticed that the canvas displaying was visually *much* slower in Nightly than Chrome.
Summary: Slower performance thatn Chrome on Voronoi polygon computations → Slower performance than Chrome on Voronoi polygon computations
Severity: normal → S3

Testing on Linux, with 10000:
Firefox: average exec time for 10000 sites = 90.8 ms (9.1 µs/site), fastest = 79 ms, slowest = 117 ms.
Chromium: average exec time for 10000 sites = 62.9 ms (6.3 µs/site), fastest = 54 ms, slowest = 126 ms.

Blocks: sm-js-perf
Severity: S3 → S4
Priority: -- → P3
You need to log in before you can comment on or make changes to this bug.