With Chrome 113, which is scheduled to launch in about three weeks, Google has announced that WebGPU, an API that gives web apps more access to your graphics card's capabilities, will be enabled by default.
A blog post claims that WebGPU offers "more than three times improvements in machine learning model inferences" and can help developers produce graphics at the same level as they can currently with much less code. The last one is really shocking; while better machine learning performance was intriguing in 2021, when the feature was added to Chrome on an experimental basis, it may now be even more useful in the era of generative AIs and sophisticated language models. There is plenty of room for cool machine learning applications that utilize your local hardware, even though services like Google's Bard and Microsoft's Bing don't really do so.
This month's release, according to Google, "serves as a building block for future updates and enhancements," promising "more advanced graphics features" and "deeper access to shader cores" along with improvements to how you actually create content that uses WebGPU, in the future.
The API has been under development for a while. Design work on it began in 2017 and has continued ever since. Additionally, it is not a Chrome-only standard; in the future, Firefox and Safari should also support it. According to Google, the company is attempting to broaden the implementation to support more operating systems, including Linux and Android.
