Speedometer 3.0 is the latest version of the browser benchmark developed in a joint effort by Google, Apple, Microsoft, and Mozilla.
Browser benchmarks were a craze for many years on the Internet, but popularity has gone down significantly in the last years.
When Google launched Chrome, it hammered home the fact that the browser was offering better performance with browser benchmarks. It was true at the time and brilliant marketing.
Mozilla, Microsoft, and other browser makers scrambled, but they had huge troubles closing the gap to Chrome.
Today, browser benchmarks are niche again. Developers and organizations may use them, but they play little role for regular Internet users. One reason for that is that browser makers have stopped using them for marketing for the most part. And the reason for that is that the gap is not as large anymore as it was 15 years ago.
Speedometer 3.0
Speedometer 3.0 is a browser benchmark by Apple, Google, Microsoft, and Mozilla. Visit the benchmark’s website in any browser and hit the “start test” button to run it.
The benchmark runs automatically from that point on. It make take a minute or two to complete the test and a score is displayed in the end. This score depends largely on the performance of the device and the browser that you use.
Speedometer 3.0 replaces the second version of the benchmark. Apple and Google collaborated in 2018 to release it to the public.
The new iteration improves “how Speedometer captures and calculates scores” and it comes with “an even wider variety of workloads”. In essence, the developers of Speedometer have updated the workloads used to test the performance of the browser to take into account changes in the last six years on the Internet.
This means that a different set of frameworks is used in the tests. New browser and JavaScript APIs are tested, and more complex computations are run as well.
You may check out Google’s post on its Chromium blog for the company’s take on the changes. Microsoft published its take on the benchmark update here.
Closing Words
Most web browsers share a common core. This is true for all Chromium-based browsers. Performance, therefore, is nearly identical when it comes to the likes of Microsoft Edge, Google Chrome, Brave, Vivaldi, or Opera. Mozilla with its Firefox web browser and Apple with its Safari browser use different engines, and results may differ more on the same machine because of that.
Unlike computer benchmarks, which may show tweaking potential, browser benchmarks come with little optimization potential. Apart from installing the latest drivers on the machine, there is little one can do to improve the performance of a browser to get better scores in benchmarks.
With that said, it is still interesting to see how your browser or system performs.
Now You: do you use benchmarks?
Thanks for another interesting article.
I’ve never run browser benchmarks before so this was something new (and easy) to try.
I use Chrome, Firefox and Slimjet browsers on a daily basis and generally avoid using Edge.
The benchmark results on my PC are interesting though:
Slimjet 9.95
Edge 9.94
Chrome 8.10
Firefox 7.53
I’m sure addins may affect the results but I think I have the same installed in each browser. I think the ranking of these results reflected the feel I have had for which the fastest one is and which was the slower. Edge ranked suprisingly well in the comparison test. That won’t encourage me to use it any more though.
I’m not a benchmark aficionado yet I’ve used previous Speedometer 2.0 only to determine the impact of given Firefox extensions on the browser’s velocity, and in some cases the impact went to some 10%…
Nice to know Speedometer has been updated to a 3.0 version. I’m off to give it a try.
I’ve just tested ‘Speedometer 3’ and, hum, the result for my browser, on this PC … doesn’t make my day. Hope it makes the reader’s laugh : I get a 2.01 …
Browser is Firefox 115.8.0 ESR, with plenty of extensions and scripts.
CPU is an old dual-core.
–
Besides personal considerations my main concern here is, not the relevancy of such a test but rather its scaling so to say : if I get a 2.01 score, if the article’s screenshot shows a 8.84 corresponding I guess to a “naked” browser but, more importantly, performed on a modern device, what are the conditions to meet to jump to a 100, 110 (and above) as Speedometer displays those as maximums : a super-computer?! In particular, given the scale, if I test again, say all extensions disabled, I’d maybe jump to 2.2 … the scale is so high that I’m wondering about the test’s relevancy when aimed to compare results on a non-powerful device.
I asked myself the same question. Maybe it is cloud-based devices or browsers running on super-computers, maybe it is just a scale that no one is able to reach at this point.