AI Weekly Malaysia

Hugging Face released @huggingface/kernels, a JavaScript library for loading and running 207 Apache-2.0 licensed WebGPU kernels directly from the HF Hub, each published as a versioned package with WGSL shader templates, correctness tests, and benchmark cases. They also launched Fleet, a browser-based benchmarking tool that crowdsources kernel performance and correctness data across real-world GPUs, letting users contribute evidence that helps identify failures and improve kernel variants.

Why: If you are building browser-based AI inference, this gives you a drop-in library of pre-optimized GPU operations (matmul, attention, quantization, convolutions) with reproducible correctness tests, potentially replacing hand-rolled WGSL shaders. The Fleet tool means you can benchmark these kernels on your own hardware before committing, and the crowdsourced evidence model helps you avoid kernels that are pathologically slow on your target devices.

Norwegian startup Orbify.eu is developing a curved 3D map rendering technique inspired by the Inception movie's dream-sequence perspective, blending near-field street-level views into overhead views at distance. The two-year-old company has a patent pending in the EU (PCT/EP2026/058725) covering the bending algorithm, shader, and camera setup, and released an interactive browser demo using the PlayCanvas engine with WebGL2/WebGPU and Gaussian splatting. The startup has initial funding from Innovation Norway and Norges forskningsråd and is still testing whether the interface actually improves navigation.

Why: The browser demo is worth a look if you build geospatial or navigation interfaces — it shows a concrete WebGL2/WebGPU + Gaussian splatting implementation in PlayCanvas that you can inspect. But the technology is pre-commercial, the patent scope is unpublished, and there's no validation yet that the warped perspective helps users navigate better, so there's nothing to adopt or change today.