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.

 

[CraptacularOne]

Umm... does your Ryzen 9 suck up 180W during Speedometer? If so, you've got some issues. My 5600X (PBO+200) peaked at 64W achieving 216.2 in Chrome private window with Steam, Plex, Citrix (hospital EMR), 10 tabs of Chrome non-private, an Excel file, and Spotify running.

It shouldn't really be concerning that the M1 (a chip designed specifically for such functions as web browsing, iTunes, and canned App Store programs) is doing well against x86 chips (designed for more than browsing Facebook and touching up photos on the toilet) in a web browser test designed by the WebKit dev team at Apple. Apple have designed the chip with their users in mind and there's a reason they haven't put the M1 in a Mac Pro.

Obviously my 5900X does not pull that much power in this, and clearly you missed the point entirely in your rush to defend your AMD CPU. My point was entirely that a mobile CPU that pulls 10-15w at max under any circumstances is at or on par with desktop variants in this browser benchmark.

Next time maybe use a little common sense before leveraging a ridiculous conclusion?

 

[Doug S]

Apple have designed the chip with their users in mind and there's a reason they haven't put the M1 in a Mac Pro.

What they do put in the Mac Pro will be even faster at Speedometer and every other web and non-web benchmark. Surely you can't possibly actually believe that Apple's SoCs and the M1 were specifically designed to run web browsers fast, and they'll have to give up web browsing performance to be fast at the sort of stuff the Mac Pro needs to be fast at?

You're foolish if you think there's any gap at all in the needs of web browsing and "canned app store programs" and whatever you think the big boy stuff you run on your x86 is. What will be your excuse when their new CPU is even faster than the M1 in every benchmark and every real world application, and is no longer handicapped by number of cores in multicore stuff?

 

[thunng8]

I haven't tested for a while, but Chrome 91 is now faster at the benchmark than Safari on my Mac mini M1. The latest version of Firefox (89) has caught up a lot as well.

Chrome 91: 241
Safari 14.1: 233
Firefox89 : 212

Note: The M1 chip was pulling less than 5W when running this benchmark.

 

[thunng8]

I haven't tested for a while, but Chrome 91 is now faster at the benchmark than Safari on my Mac mini M1. The latest version of Firefox (89) has caught up a lot as well.

Chrome 91: 241
Safari 14.1: 233
Firefox89 : 212

Note: The M1 chip was pulling less than 5W when running this benchmark.

I ran it again which any other apps open and got 246 on Chrome 91

 

[amd6502]

Ok, now this is more like a real browser. Now I get a decent score of 47 using Firefox ESR (78) on my linix laptop (2500u @ 15w) and Cinnamon window manager.

I get a score of under 6 with Chrome on my Kindle Fire, which uses a low powered acorn quadcore. I suspect the cores might be in-order execution. The IPC feels lower than anything I've seen. Even a P4 will probably run circles around it.

It is usually not suitable for most browsing, even with one or barely any tabs open. It's best to have a physical book or other reading material handy while the browser computes and you wait for the next frame. Even scrolling will hang for seconds often.

 

[.vodka]

3900X + PBO fmax enhancer, 3800MHz 16-19-19-38 1T+gdm tuned memory
R9 290, 21.3.1
W10 20H2 19042.867

Speedometer 2.0
-----------------------------
Chrome 89: 154.1 ± 1.1
Chromium Edge: 153 ± 1.6
Firefox 87 : 146 ± 2.0

Octane 2.0
-----------------------------
Chrome 89: 57077
Chromium Edge: 56687
Firefox 87 : 26232, lol something is wrong here

Mozilla Kraken 1.1
-----------------------------
Chrome 89: 670.6ms ± 0.3%
Chromium Edge: 677.8ms ± 0.2%
Firefox 87 : 842.7ms ± 1.0%

WebXPRT 3
-----------------------------
Chrome 89: 270


All with no extensions enabled. Zen3 is stupid fast in these relative to Zen2.

Chrome 91.0.4472.101 64b:
Speedometer 2.0: 173.9 ± 1.7
Mozilla Kraken 1.1: 656.6ms +/- 0.4%
Octane 2.0: 57772

Firefox 89 64b:
Speedometer 2.0: 155 ± 1.7
Mozilla Kraken 1.1: 856.8ms +/- 0.8%
Octane 2.0: 25985

Google's latest improvements make for quite the speedup on Zen2 too. Mozilla was also at it, at least as measured by Speedometer.

 

[amrnuke]

What they do put in the Mac Pro will be even faster at Speedometer and every other web and non-web benchmark.

Yes, but what they put in the Mac Pro will also use more power. It'll probably still be more power efficient than the comparable AMD or Intel chip though.

Surely you can't possibly actually believe that Apple's SoCs and the M1 were specifically designed to run web browsers fast

I do not believe that, and I did not say that. The A14 and M1's cores were designed to run with aplomb the common tasks that their target audience uses the most - like browsing, yes, but also image/video editing, light gaming, and so on. If the Firestorm and Icestorm cores ran browsers at half the speed they currently do, Apple would have fixed that because the end-user experience would have been crappy.

and they'll have to give up web browsing performance to be fast at the sort of stuff the Mac Pro needs to be fast at?

Didn't say that. Why would you even think I believe that? Did Zen3 give up web browsing performance to be good at rendering and media editing and gaming?

You're foolish if you think there's any gap at all in the needs of web browsing and "canned app store programs" and whatever you think the big boy stuff you run on your x86 is.

Didn't say that. But that's an interesting take. So there's no gap between the needs of a web browser and, say, using JMP to analyze management patterns in a medical practice, or batch processing the 100+ pictures I took last week?

What will be your excuse when their new CPU is even faster than the M1 in every benchmark and every real world application, and is no longer handicapped by number of cores in multicore stuff?

My "excuse" will be that this is entirely expected.

 

[Det0x]

*edit*
will post later

 

[Det0x]

Chrome version 91.0.4472.106

 

[gplracer]

Edge
I5-4670K at 3.4ghz
8 gigs of ram

 

[plopke]

Ryzen 1700 no OC , gtx 970 , B350 , 16GB 2400 17-17-17

Firefox build 89.0.1 64-bit : 75.6
Edge chromium Version 91.0.864.54 (Official build) (64-bit) : 102
Google chrome Version 91.0.4472.114 (Official Build) (64-bit) : 102

 

[jihe]

Ryzen 1200AF @4ghz, Firefox 89.0.1: 103

 

[tamz_msc]

i7-11370H with 16 GB DDR4-3200@JEDEC timings - 25W PL1, 51W PL2. Edge 91.

 

[Mir96TA]

Here is my i3-8100

 

[ptthomas]

Mines isn't displaying anything

 

[JoeRambo]

i7-11370H with 16 GB DDR4-3200@JEDEC timings - 25W PL1, 51W PL2. Edge 91.

Impressive score for mobile CPU, very near peak 10900K performance on mobile platform.

 

[thunng8]

Ran it again, this time making sure no other applications were running. Also ran in incognito this time.

I think a new record. 274.

The M1 and Chrome is getting very fast at this benchmark.

Chrome 91.0.4472.114

I also checked power usage of the CPU - still only peaking at 5W.

 

[Heartbreaker]

40.7 with Vivaldi Browser.

I was expecting worse, since my computer is 2008 vintage C2Q and 8800GT GPU.

 

[john3850]

My score went up by 11 to 199 after a windows update strange.
The first time I ran this my score was 146 now a year later I can get 201 on the same pc.

 

[Magic Carpet]

40.7 with Vivaldi Browser.

I was expecting worse, since my computer is 2008 vintage C2Q and 8800GT GPU.

Chrome should be faster!

Old but never forgotten: T7600 + GMA950 + Win 8.1 Pro w/ Media Center 32-bit + Chrome v91. Not bad for a 14 year old laptop 🤣

 

[scannall]

111.1 on a bonestock 2600X, with stock ram settings running Firefox.

 

[Magic Carpet]

I don't really understand how those tests were done. Seem a bit off, imo.

Looks like the newest version of edge is much faster now. Scores are much closer now between Edge and chrome compared to previous versions.

Chrome 92.0.4515.131 seems to be consistently around 10% faster than Edge on my Windows 10 Pro 21H1 system:

 

[burninatortech4]

Edge 92.0.902.73 - 5800X Stock

 

[thunng8]

Safari 15 got a nice boost compared to Safari 14 on the Mac mini m1

257.9

(Safari 14 was 233)

Still below that of Chrome though (274)

 

[Asterox]

View attachment 48763

I don't really understand how those tests were done. Seem a bit off, imo.



Chrome 92.0.4515.131 seems to be consistently around 10% faster than Edge on my Windows 10 Pro 21H1 system:

View attachment 48764

Well, in short this Webbench is useless garbage for use out of boredom.

Anandtech scores for R5 4650G is 98, and hm R7 5800X received 141.

Ok, but what CPU do i use does anyone have some idea.

- browser Google Chrome