Google, Apple, Microsoft and Mozilla release Speedometer 3.0 to benchmark browser performance

DragonSlayer101

Posts: 405   +2
Staff
What just happened? The four largest browser developers in the world – Google, Apple, Microsoft, and Mozilla – have collaborated to develop the Speedometer 3.0 benchmarking tool. For the uninitiated, Speedometer measures key performance metrics of browsers and the web app is often used to determine which is the fastest browsing software on the planet.

Speedometer 3.0 is the third major version of the popular benchmarking tool that measures app responsiveness by simulating real-world user interactions on web pages. The latest version is a solid upgrade over Speedometer v2.0 that was released in 2018 and includes many new kinds of sub-tests, contemporary frameworks, and more. The second-generation app was a much-needed refresh to the original version released in 2014, but had begun to show its age in recent years.

Non-profit BrowserBench highlights that this is the first time any major browser benchmark has been developed through a cross-industry collaboration supported by each major browser engine, including Blink/V8, Gecko/SpiderMonkey, and WebKit/JavaScriptCore. They also claim that Speedometer 3.0 marks a major step forward in web browser performance testing by introducing a "better way of measuring performance (through) a more representative set of tests that reflect the modern Web."

Speedometer 3.0 adds many new tests to better gauge browser performance in some key scenarios. Some of the new tests simulate rendering canvas and SVG charts, such as React Stockcharts, Chart.js, Perf Dashboard, and Observable Plot. It also takes into account code editing (CodeMirror), WYSIWYG editing (TipTap), and reading news sites (Next.js and Nuxt.js).

v3.0 also improved the TodoMVC tests by updating the code to adapt to the most common versions of the most popular frameworks based on data from the HTTP Archive. The included frameworks and libraries in the new version include Angular, Backbone, jQuery, Lit, Preact, React, React+Redux, Svelte, and Vue, along with vanilla JavaScript implementations targeting ES5 and ES6, and a Web Components version. The blog post also details many other complex tests that the developers believe will help better evaluate browser performance on the modern web.

Over the years, Speedometer has become the default tool for browser developers to drive performance optimizations as websites continue to get more complex and data-heavy. While v1.0 was created by Apple's WebKit team, the second-generation software was a collaboration between Google and Apple. This time around, Microsoft and Mozilla also got involved in the process.

Permalink to story.

 
Does Intel still win easily? They dominated Speedometer 2.0

Even i5-13600K beat 7950X

Real question, does this comparison also accounts for which browser was used? Because I would think that some browser would run better/worse on different CPUs, like Edge runs better on 7950x, Chrome runs better on i5-14900K... It's been a while since I looked at browser benchmarks.

I tried it on my PC with Vivaldi that's my default browser, I'm thinking about benchmarking Chrome and Edge just for comparison, under same hardware configuration.
 
This is a benchmark for browser speed, why are the commentors asking about the fastest CPUs?

Also, in a list of 3 For Profit companies that want your data vs 1 Nonprofit that doesn't want your data, pick the one that doesn't want your data.

My point is that if browsers ranks in same positions no matter the hardware configuration. I don't know if browsers benchmarks take that into account or just do the benchmark under just one setup. That would give a different insight about browser performance... I'm a AMD user, I could give X browser preference over another if it runs better on AMD.
 
Lets see if Techpowerup includes Speedometer 3.0 in the 14900KS testing later this week.

browser-speedometer.png
 
I just tried the link in your story and got zero results.
All it did was fast flash through a bunch of instructions then gave me a blank screen to work with.
Not very impressed.👎☹️
 
Back