Chrome renderer uses over 3x CPU vs. Safari renderer on 163.com |
|||||
Issue descriptionChrome Version: 60.0.3089.0 OS: macOS 10.12 On my 12 core Mac Pro tower, with Turbo Boost disabled and processor frequency pegged at max, I have the following for CPU utilization over 30s on the site 163.com: Chromium 60.0.3089.0 Browser 0.9% Renderer 17.9% GPU 0.9% Window Server 3.7% Safari 10.1 Browser 0.2% Renderer 5.4% Window Server 1.7% Window Server baseline of 1.0%. Chrome's non-renderer processes are doing relatively little work, so we're basically comparing Chrome renderer (17.9%) against Safari's renderer (5.4%). WebKit is able to do equivalent work with 1/3 the CPU. The 163.com site is archived at https://drive.google.com/open?id=0BwcLL5PHtZQHRWhqaVllQlRDVlk Chrome Tracing trace: https://drive.google.com/open?id=0BwcLL5PHtZQHTlZMWE1PSHdSV2s Instruments 8.2.1 (macOS tool) trace of the renderer process: https://drive.google.com/open?id=0BwcLL5PHtZQHRXNIei1SMy00cDQ
,
May 10 2017
Is this a recent regression? What does M-58 do?
,
May 10 2017
Using bisect-builds to provide versions of Chromium to benchmark, I see: 58.0.3029.110 Renderer - 16.2% 60.0.3089.0 Renderer - 19.0% This suggests that things have gotten worse recently (by almost 20%), but they were not great in M58 (still around 3x of Safari). Back to schenney@ so that you see this comment and perhaps retriage.
,
May 11 2018
This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue. Sorry for the inconvenience if the bug really should have been left as Available. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
May 11 2018
|
|||||
►
Sign in to add a comment |
|||||
Comment 1 by e...@chromium.org
, May 10 2017