The Power of the GPU in the Palm of your Browser, with WebGPU

In this talk, we will dive into the magical world of WebGPU, the next-generation graphics API that is set to revolutionize how we create and interact with 3D graphics on the web. WebGPU is the successor to WebGL, designed to provide more direct and efficient access to modern GPU hardware, enabling high-performance graphics and computation right in your browser. Imagine having a software like Photoshop with all of it's features just available to run in any browser on any device!

Not only graphics, but anything that involves compute. With the help of compute shaders ( https://developer.chrome.com/docs/capabilities/web-apis/gpu-compute#shader_programming ), we are able to to complex GPU tasks right in the browser engine. For this talk I will be focusing more on the graphical compute stack.

WebGPU has been supported in development by numerous organizations such as Apple, Google, Intel, Mozilla, and Microsoft. Currently, all major web browsers have implemented or are in the process of implementing WebGPU. Before end of the year every browser will have it in stable releases.

1. First, we'll start with the introduction of WebGPU and why is it better than older web graphical stacks like WebGL.
2. We'll dive into what problems the current web ecosystem has when it comes to compute and graphics and how WebGPU solves them.
3. Introduction and Demo of a project that uses WebGPU: https://github.com/rerun-io/rerun ; Rerun is an open source tool that helps you record, store, and visualize data that changes over time. It's useful in areas like robotics, simulations, and finance to check, fix, and understand data. All in the web/in your browser!