Announcement

Collapse
No announcement yet.

Chrome 89 vs. Firefox 86 Performance Benchmarks On AMD Ryzen + Ubuntu Linux

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Chrome 89 vs. Firefox 86 Performance Benchmarks On AMD Ryzen + Ubuntu Linux

    Phoronix: Chrome 89 vs. Firefox 86 Performance Benchmarks On AMD Ryzen + Ubuntu Linux

    Given this week's launch of Chrome 89 and the recent Firefox 86 debut, here are some quick benchmarks for those curious about the current performance when using Ubuntu Linux with a AMD Ryzen 9 5900X and Radeon graphics...

    http://www.phoronix.com/scan.php?pag...-89-Firefox-86

  • #2
    Meanwhile, Chromium is already on v91.

    Also, Nouveau users and those on older Wayland compositors like Kwin 5.14 and Gnome 3.30 should skip all Chromium / all Chromium-based browsers on v90 and go straight to v91, because something is broken in v90. Tested and confirmed in both Chrome Beta, Chrome Dev, Chromium 90 and Edge Dev for Linux.

    And I can't be assed to tell you what got broken in v90. Go dig through the Chromium bug reports and mailing lists on your own to find out. Because that's what Linux 'elites' love to do; read and file bug reports.

    Comment


    • #3
      Still not enough reasons to use spyware from data stealing company called google.

      Comment


      • #4
        Was this tested with webrender enabled on Firefox? Any performance tweaks? Was Chrome using the gpu for everything?
        Showing at least chrome://gpu and the equivalent in Firefox would have been nice.

        Comment


        • #5
          You should have tested Edge as well

          Comment


          • #6
            Everybody should learn their lesson and move to Firefox seeing what happened with Chromium sync...
            ## VGA ##
            AMD: X1950XTX, HD3870, HD5870
            Intel: GMA45, HD3000 (Core i5 2500K)

            Comment


            • #7
              It's like comparing a 600 bhp Dacia to a 200 bhp Mercedes.

              Chrome is completely stripped down (of features) to the max to deliver faster performance. It's all about the engine. Kind of a NASCAR car as well. But there's basically nothing you can do or tweak with it. And your workfklow is taking a hit by the time you make up for the features it doesn't have (the lack of a favorite button is a quick example) or to deal with the crap (no global setting to switch off video autoplay).

              With Firefox, you might not drive as fast, although it's not very noticeable anyway in most real-life scenarios, but damn that seat is confortable, oh and that little button to activate AC or to roll the window down or the panoramic roof open, it's so cozy. Try to do that on your Chrome Dacia.

              Originally posted by szymon_g View Post
              You should have tested Edge as well
              Indeed, it is a better Chrome than Chrome itself.
              Last edited by Mez'; 04 March 2021, 08:07 AM.

              Comment


              • #8
                Originally posted by juxuanu View Post
                Was this tested with webrender enabled on Firefox? Any performance tweaks? Was Chrome using the gpu for everything?
                Showing at least chrome://gpu and the equivalent in Firefox would have been nice.
                WebRender has nothing to do with raw JavaScript performance.

                Originally posted by darkbasic View Post
                Everybody should learn their lesson and move to Firefox seeing what happened with Chromium sync...
                Sorry, what? Been using Chrome for ages, hasn't been affected by any changes.

                If you insist on using a compatible web browser - that's your choice but you should not expect Google to cater to you.

                Comment


                • #9
                  Originally posted by juxuanu View Post
                  Was this tested with webrender enabled on Firefox? Any performance tweaks? Was Chrome using the gpu for everything?
                  Showing at least chrome://gpu and the equivalent in Firefox would have been nice.
                  Webrender should be enabled by default on this setup (https://wiki.mozilla.org/Platform/GFX/WebRender_Where). What I would find interesting would be testing EGL backend, as benchmarks using WebGL should get quite a bit faster. I.e. `MOZ_X11_EGL=1` in a X11 session or `MOZ_ENABLE_WAYLAND=1` in a Wayland session.

                  Some of the benchmarks are AFAIK not really useful, for example octane. See https://www.infoq.com/news/2017/04/o...hmark-retired/ :
                  By 2015, however, most JavaScript implementations had implemented the compiler optimizations needed to achieve high scores on Octane. In addition, we began to notice that JavaScript optimizations which eked out higher Octane scores often had a detrimental effect on real-world scenarios.
                  Speedometer is the one I personally consider most interesting of those shown here. What I miss is a benchmark that simulates "average" browsing behaviour such as loading a website for the first time etc. AFAIK the FF JS engine, because of its different architecture, has an edge there as it is better at ignoring unused JS. But I don't have numbers apart from FF claims that they do better in some of these scenarios.

                  Comment


                  • #10
                    birdie darkbasic said chromium not chrome.

                    Comment

                    Working...
                    X