5 ms·

Worth looking at this resource, which seems to suggest that Chrome still leads in some of the benchmarks, even in memory usage: https://www.phoronix.com/scan.p

Worth looking at this resource, which seems to suggest that Chrome still leads in some of the benchmarks, even in memory usage: https://www.phoronix.com/scan.php?page=article&item=firefox-quantum-bench&num=1 https://www.phoronix.com/scan.php?page=article&item=firefox-... Just one study among many, but, worth keeping in mind. Firefox is making strides, but Mozilla still has work to do if they want to stop losing market share. I say this as a Firefox user.

amaranth 9y ago
Most of those benchmarks are (directly or indirectly) benchmarks of the JavaScript engine. The few that aren't are instead benchmarking the <canvas> implementation or the compositor. The main speedups in Firefox 57 are in styling and the DOM which are more of a "feels faster" thing than something you can easily benchmark in real-world scenarios other than "how long does this page take to load?"
nnethercote 9y ago
Here's a blog post from this week about memory usage: http://www.erahm.org/2017/09/25/firefox-memory-usage-in-the-quantum-era/ http://www.erahm.org/2017/09/25/firefox-memory-usage-in-the-... That's a comparison on real websites rather than browser benchmarks.
482794793792894 9y ago
For now, it's already been enough to stop the losing of market share. The introduction of multiprocess in Firefox 48 (2nd August 2016) seems to have made the difference: http://netmarketshare.com/report.aspx?qprid=1&qpaf=&qpcustom=Firefox&qpcustomb=0&qpsp=200&qpnp=24&qptimeframe=M http://netmarketshare.com/report.aspx?qprid=1&qpaf=&qpcustom... But yeah, while the technical crowd sees Mozilla as favorable at roughly equal performance, the vast majority of unknowing users are probably more influenced by the world's most popular webpage, video portal, e-mail service, operating system etc. all nudging them to use Chrome. It'll be interesting to see, if Firefox 57 can generate another major bump in that user graph like it happened with Firefox 48, or if 48 was at sort of the sweet spot where Firefox became acceptable again for most of the tech crowd and now we'd be digging into the non-tech crowd, which is much harder to pull away from what they're habitually using.

amaranth 9y ago

Most of those benchmarks are (directly or indirectly) benchmarks of the JavaScript engine. The few that aren't are instead benchmarking the <canvas> implementation or the compositor. The main speedups in Firefox 57 are in styling and the DOM which are more of a "feels faster" thing than something you can easily benchmark in real-world scenarios other than "how long does this page take to load?"

nnethercote 9y ago

Here's a blog post from this week about memory usage: http://www.erahm.org/2017/09/25/firefox-memory-usage-in-the-quantum-era/ http://www.erahm.org/2017/09/25/firefox-memory-usage-in-the-... That's a comparison on real websites rather than browser benchmarks.

482794793792894 9y ago

For now, it's already been enough to stop the losing of market share. The introduction of multiprocess in Firefox 48 (2nd August 2016) seems to have made the difference: http://netmarketshare.com/report.aspx?qprid=1&qpaf=&qpcustom=Firefox&qpcustomb=0&qpsp=200&qpnp=24&qptimeframe=M http://netmarketshare.com/report.aspx?qprid=1&qpaf=&qpcustom... But yeah, while the technical crowd sees Mozilla as favorable at roughly equal performance, the vast majority of unknowing users are probably more influenced by the world's most popular webpage, video portal, e-mail service, operating system etc. all nudging them to use Chrome. It'll be interesting to see, if Firefox 57 can generate another major bump in that user graph like it happened with Firefox 48, or if 48 was at sort of the sweet spot where Firefox became acceptable again for most of the tech crowd and now we'd be digging into the non-tech crowd, which is much harder to pull away from what they're habitually using.