GPU Programmer - Remote

We are looking for a remote GPU programmer to build and fine‑tune GPU‑driven components for artificial‑intelligence and large‑language‑model workloads. The day‑to‑day work involves creating and improving GPU code with CUDA, WebGPU or GLSL, measuring and speeding up kernels and shaders, writing the host‑side C++ glue, and turning performance problems into optimized solutions. Applicants must have solid experience programming GPUs, especially NVIDIA devices, and be comfortable with CUDA, WebGPU or GLSL as well as advanced C++ development. A background in graphics rendering, machine‑learning acceleration, scientific computing or high‑performance computing is also required, together with a deep grasp of GPU architecture and optimisation techniques. The position is offered on a contract basis and can be performed from any location with reliable internet. What the role asks for: - Proven GPU programming experience, especially NVIDIA GPUs - Proficiency in CUDA, WebGPU or GLSL - Advanced C++ development skills - Background in graphics, ML acceleration, scientific computing or HPC - Deep understanding of GPU architecture and performance optimisation