All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=830775

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=9008faa53cf552b3adebf7110d35399cba29c83d55defc9e5d16bf0c917e4c10


Bot(s) for this bug's original alert(s):

chromium-rel-win10

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14d6041cc40000

📍 Found significant differences after each of 3 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/14d6041cc40000

Move Buffer Subject/Observer to front end. by jmadill@chromium.org
https://chromium.googlesource.com/angle/angle/+/0946393dfd115f5d74abfc8ad4b8ef7b4d363503

SQLitePersistentCookieStore: Don't leak on major errors in Commit by morlovich@chromium.org
https://chromium.googlesource.com/chromium/src/+/0f0ca6ab9a3d88d72f8a0b13d5cf8f236f846cfa

Make SpeechRecognitionBrowserTest.OneShotRecognition use the test server by mmenke@chromium.org
https://chromium.googlesource.com/chromium/src/+/b5ce7dd19171acf855438692d735ff692c7ebc47

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

A change in a browser test could not have caused a regression in something other than a browser test.  Clearly either the bisect's wrong, or there's something else going on here.

The angle change seems like the likely suspect here given it's the 3D graphics library. 

Move Buffer Subject/Observer to front end. by jmadill@chromium.org
https://chromium.googlesource.com/angle/angle/+/0946393dfd115f5d74abfc8ad4b8ef7b4d363503

This could be the issue, I'm working on fixing a regression for that in  issue 829906 . Could dupe it into that issue even.