Google Chrome is a closed-source web browser developed by Google
Google unveiled recently through a blog post on new maglev compiler which aims to quickly generate high-performance native code for actively used JavaScript code.
This new JIT compiler, Maglev, stands out for being a mid-level Chrome compiler which "can rapidly generate high-performance machine code for all relevant functions in the first hundredth of a second."
Since the beginning of Chrome, benchmarking has been a key way we drive performance optimizations that benefit users. The most relevant web benchmarks today are Speedometer, MotionMark, and Jetstream. Over the past year, Chrome has invested in optimization against these specific benchmarks and just achieved our highest scores on all three. These gains were achieved through a combination of large projects and small improvements. In today's The Fast and the Curious post, we want to share some of the ways we're driving these improvements to Chrome.
In their blog post Google mentions that the inclusion of Maglev pIt allowed it to speed up the Jetstream performance test by 7,5% and the Speedometer test by 5%.
In addition, Chrome's overall performance dynamics are mentioned:
- In the Speedometer test, which focuses on evaluating the browser's responsiveness to websites and measuring the execution speed of popular JavaScript libraries, Chrome score improved from 330 to 491. In addition to switching to Maglev, the tests took into account other optimizations made in releases over the past year (since release 101), for example, function call optimizations in the JavaScript engine.
- In the Jetstream test, designed to test work with advanced JavaScript and WebAssembly web applications, the use of Maglev achieved a score of 330 points (an improvement of 7,5%).
- In the test MotionMark, which tests the capability of the graphics subsystem of a browser to generate information at a high frame rate, performance improved three times from last year. Since the beginning of the year, developers have proposed more than 20 optimizations that make working with graphics faster in Chrome, half of which are already included in the stable codebase.
For example, canvas performance has been improved, optimizations based on code profiling have been enabled, GPU-side task scheduling has been improved, layering (compositing) performance has been improved, a new MSAA (Multisample Anti-Aliasing) dynamic smoothing algorithm has been implemented and 2D canvas rasterization removed in separate processes to parallelize operations.
Regarding the implementation of Maglev, it is mentioned that this will be rolled out gradually for Chrome 114 users and if you are interested in being able to know more about it, you can consult the Google blog post In the following link.
In addition to the Maglev implementation, we also It is worth noting that Google announced that it has stopped using the Dialer and Messaging applications provided by the repository AOSP (Android Open Source Project), which develops the open source Android platform.
It is announced that in the future, these applications that provide an interface for making calls and working with SMS will be removed from the AOSP repository, which will require the use of external applications when building builds based on AOSP to implement this functionality.
At the same time, deprecated applications were generally not used in practice and were mainly regarded as prototypes worksheet with examples of how to work with calls and text messages. Phone carrier-provided firmware and AOSP-based builds such as LineageOS used their own implementations of the Dialer and Messaging apps.
The end of support for applications in AOSP was announced in The next message.