Question How fast is your browser?

[Magic Carpet]

This benchmark simulates user actions for adding, completing, and removing to-do items using multiple examples in TodoMVC. Each example in TodoMVC implements the same todo application using DOM APIs in different ways. Some call DOM APIs directly from ECMAScript 5 (ES5), ECMASCript 2015 (ES6), ES6 transpiled to ES5, and Elm transpiled to ES5. Others use one of eleven popular JavaScript frameworks: React, React with Redux, Ember.js, Backbone.js, AngularJS, (new) Angular, Vue.js, jQuery, Preact, Inferno, and Flight. Many of these frameworks are used on the most popular websites in the world, such as Facebook and Twitter. The performance of these types of operations depends on the speed of the DOM APIs, the JavaScript engine, CSS style resolution, layout, and other technologies.

Speedometer 2.0

Core M-5Y10c @ 52.02

Images aren't necessary, but please state your cpu speed. The web browser of your choice. Mine is Edge 83. Thank you.

 

[amrnuke]

System: 5600X, PBO +200, minor CO offsets 0 (best core) to -10 (worst 4 cores), 3666 CL 16 RAM, 2.5" SATA SSD, Radeon 5700 at stock
All run in private mode (no extensions)

Chrome: 182.4
Edge: 179.9
Firefox: 157.0
Brave: 155.8

Looking @Det0x -- it's gotta be my FCLK and RAM timings holding me back, no way this has anything to do with you having a 5950X and me having a 5600X

 

[Midwayman]

192 with my iphone 12pro.

 

[Det0x]

System: 5600X, PBO +200, minor CO offsets 0 (best core) to -10 (worst 4 cores), 3666 CL 16 RAM, 2.5" SATA SSD, Radeon 5700 at stock
All run in private mode (no extensions)

Chrome: 182.4
Edge: 179.9
Firefox: 157.0
Brave: 155.8

Looking @Det0x -- it's gotta be my FCLK and RAM timings holding me back, no way this has anything to do with you having a 5950X and me having a 5600X

What do you get in other singlethread benchmarks ?
over 700 in cpu-z bench?

 

[amrnuke]

What do you get in other singlethread benchmarks ?
over 700 in cpu-z bench?

mid-660s (662-666 range)
CB23 ST 1609
CB20 ST 623
GB5 ST 1676

After looking into it, my browser performs much better on the first new-instance run. Subsequent repeated runs without closing and restarting the browser are slower. On the first new-instance run, there are noticeable hitches at test 14 and 17, but the rest goes quickly. On subsequent runs, everything goes very smoothly but scores are lower. I am not sure if that hitch is allowing the CPU to cool down, though in that segment of the test it's primarily very lightly threaded.

 

[Avalon]

Got a 165.2 with a stock 5600X using the current version of Chrome.

 

[john3850]

173.8 is the best I can do with 8 cores on the Edge Dev.

 

[Nothingness]

220 +/- 8.9 Safari 14.0.2 on Apple MBP with M1 @ 3.2 GHz

 

[amrnuke]

220 +/- 8.9 Safari 14.0.2 on Apple MBP with M1 @ 3.2 GHz

This is a bafflingly high score. There is no doubt that vertical integration of hardware and software has great potential to create massive performance gains for end-users.
What does the M1 do with native Chrome or Firefox for you?

 

[Midwayman]

This is a bafflingly high score. There is no doubt that vertical integration of hardware and software has great potential to create massive performance gains for end-users.
What does the M1 do with native Chrome or Firefox for you?

Not surprising after my iphone score.I think they both use safari to render anyways.

 

[amrnuke]

Not surprising after my iphone score.I think they both use safari to render anyways.

Unless something changed, I thought Chrome on macOS used Blink and not WebKit. On iOS you are right, Chrome uses WebKit.

 

[Midwayman]

Unless something changed, I thought Chrome on macOS used Blink and not WebKit. On iOS you are right, Chrome uses WebKit.

right, iOS. Though with the move to arm who knows?

 

[amrnuke]

right, iOS. Though with the move to arm who knows?

No, I mean Chrome on Big Sur on Arm does use Blink, unless something changed since the native app was released in November.

Anyway, what I was saying is, I'm curious if M1-native Chrome / Firefox are as fast as Safari.

 

[sallymander]

No, I mean Chrome on Big Sur on Arm does use Blink, unless something changed since the native app was released in November.

Anyway, what I was saying is, I'm curious if M1-native Chrome / Firefox are as fast as Safari.

I got a score of 210 +- 1.4 on my M1 Mac Mini

Chrome Version 87.0.4280.141 (Official Build) (arm64)

 

[moinmoin]

I got a score of 210 +- 1.4 on my M1 Mac Mini

Chrome Version 87.0.4280.141 (Official Build) (arm64)

Thanks! Nice confirmation of both M1 being fast and Apple's claim of Safari being the fastest browser appearing to be correct.

 

[amrnuke]

I got a score of 210 +- 1.4 on my M1 Mac Mini

Chrome Version 87.0.4280.141 (Official Build) (arm64)

Sounds to be measurably off Safari results, right?

Edit: Just perused Arstechnica's article on the native Chrome app, it is indeed 5-10% off Safari's scores.

Kind of odd now that there are no mainstream WebKit / JSCore-based browsers for Windows.

 

[moinmoin]

Kind of odd now that there are no mainstream WebKit / JSCore-based browsers for Windows.

Google decided to fork it to create Blink since they had different ideas regarding sandboxing. They are open source, so if anybody wants to create a browser based on Safari's rendering and JavaScript engine that's still possible.

 

[amrnuke]

Google decided to fork it to create Blink since they had different ideas regarding sandboxing. They are open source, so if anybody wants to create a browser based on Safari's rendering and JavaScript engine that's still possible.

Would be interesting to see if the speed advantage over Blink/V8 on macOS would translate to Windows.

 

[moinmoin]

Would be interesting to see if the speed advantage over Blink/V8 on macOS would translate to Windows.

It should. However WebKit/JSCore is only the innermost part of a browser, everything else (transmission and caching of data, handling of cookies and permissions etc. pp.) would need to be implemented first. Which is why essentially all the derivative browsers build upon Chromium instead which is a complete cross-platform browser available as open source. (Also, I'm not sure if that changed since, but one of the original causes of Google forking WebKit was Apple constantly unilaterally introducing MacOS/iOS-isms into the code that first needed to be ported to work on other platforms.)

 

[amrnuke]

It should. However WebKit/JSCore is only the innermost part of a browser, everything else (transmission and caching of data, handling of cookies and permissions etc. pp.) would need to be implemented first. Which is why essentially all the derivative browsers build upon Chromium instead which is a complete cross-platform browser available as open source. (Also, I'm not sure if that changed since, but one of the original causes of Google forking WebKit was Apple constantly unilaterally introducing MacOS/iOS-isms into the code that first needed to be ported to work on other platforms.)

Very true, though it appears the Blink/V8 based browsers I've tested (Chrome, Edge, Brave) all achieve statistically similar scores when extensions are off. Firefox is a laggard on pretty much every system I've tested.

 

[tamz_msc]

i5-1135G7 4.2GHz Turbo 2.4 GHz base

With Edge(chromium) latest:



With Firefox Developer latest:



Huge disparity.

 

[moinmoin]

Very true, though it appears the Blink/V8 based browsers I've tested (Chrome, Edge, Brave) all achieve statistically similar scores when extensions are off.

Well yes, they are all Chromium based so the deviation should be small.

To get on topic again, let me check my browsers on my lowly N3700 Linux netbook...
Opera 73.0.3856: 12.2 ± 0.77
Chromium 87.0.4280: 13.8 ± 0.61
Firefox 84.0.2: 10.9 ± 0.62
Konqueror 20.08.3: 13.7 ± 0.88
Epiphany 3.38.2: 12.2 ± 1.0

 

[amrnuke]

Well yes, they are all Chromium based so the deviation should be small.

I guess I was expecting some Microsoft "special sauce" to help Edge out more.

To get on topic again, let me check my browsers on my lowly N3700 Linux netbook...
Opera 73.0.3856: 12.2 ± 0.77
Chromium 87.0.4280: 13.8 ± 0.61
Firefox 84.0.2: 10.9 ± 0.62
Konqueror 20.08.3: 13.7 ± 0.88
Epiphany 3.38.2: 12.2 ± 1.0

Did you forget to turn it on?

 

[moinmoin]

Did you forget to turn it on?

LOL! I think the low numbers are down to the lack of decent hardware acceleration under Linux (though WebGL works fine at least).

 

[teejee]

My iphone 12 Pro Max gets the second highest score in this thread - 197, only beaten by M1 Mac

 

[Midwayman]

My iphone 12 Pro Max gets the second highest score in this thread - 197, only beaten by M1 Mac

Yah. My 12 pro is insanely fast, but my web experience is still better on my 5900x just due to being able to run a proper ad block.