Compute Architect

NVIDIA, the world leader in accelerated computing, is seeking a Compute Architect to join their team. This role sits at the crucial intersection of compiler technology and computer architectures for deep learning. The position involves developing and optimizing MLIR-based compiler infrastructure that powers NVIDIA's deep learning libraries and influences future GPU architectures. The ideal candidate will have strong expertise in compiler development, particularly with MLIR or LLVM frameworks, and deep understanding of computer architecture and performance optimization. They will work on cutting-edge projects involving deep learning algorithms, especially matrix multiplication and convolution. This role offers the opportunity to make significant impact in a fast-moving, technology-focused company, working with world-class teams at the forefront of AI and hardware innovation. The position requires collaboration with hardware architects and software teams, making it perfect for someone who thrives in a cross-functional environment and wants to shape the future of machine learning infrastructure. Located in Shanghai, this role combines technical depth with strategic influence on NVIDIA's compiler technology direction.