Spanning a wide array of platforms – from palm-sized mobile devices to vast data centers – GPUs have become a solid foundation for innovation, enabling technology that truly feels like magic. But there are no arcane spellweavers inhabiting our computational devices. Only cleverly authored bits and bytes running on specialized architecture.

However, while GPUs are everywhere, fully leveraging their capabilities remains a complex puzzle for many developers. Navigating the vendor-specific API landscape and mastering parallel programming is a daunting task. Nevertheless, with the advent of the upcoming WebGPU API bringing high-performance computing to the web, and the success of modern APIs such as Vulkan and CUDA, the era of computationally demanding products is already upon us. Making it the perfect time to learn the fundamentals of GPU programming.

Learning objectives:
- Learn the fundamentals of parallel programming and how to avoid common pitfalls.
- Understand the general architecture of modern GPUs.
- Understand the role of Vulkan as a modern GPU API
- Understand the role of WebGPU as a low-level API that exposes modern GPU APIs directly to the browser.
- Learn how to set up, optimize, and initiate GPU computations in WebGPU.

By the end of this talk, attendees will have a foundational understanding of GPU architectures, the role of WebGPU and Vulkan, and the basics of writing GPU-accelerated code. Join the quest to unravel the potential of GPUs – no magic wand required.