Overview As a core contributor to CUTLASS, you will design high-performance Pythonic abstractions and robust compilation workflows to bring GPU kernels from Python to efficient machine code. You will shape the developer experience for GPU programming, deliver open-source tooling and documentation, and help scale CUTLASS DSL to empower AI researchers. You'll bridge low-level primitives with high-level productivity, enabling fast, reliable deployment of kernels in Python environments. This role offers hands-on impact on the Python exposure of CUDA-based math primitives and compiler-focused tooling. Compensation / Benefitsequitybenefits ResponsibilitiesDesign user-focused APIs with a native feel for Python-user developersDevelop robust compilation infrastructure including AST transformations and JIT-friendly executionOptimize developer experience with debugging tools, profilers, and validation methodologiesBuild production-grade delivery infrastructure for the open-source community (packaging, wheels, conda) and documentation/testingCollaborate on delivering high-performance, Python-extended GPU programming workflows Key requirementsMS or PhD in CS, EE, or related field (or equivalent experience)3+ years of relevant experienceStrong proficiency in Python and C++, especially Python extensions and FFIExperience in library or framework development with intuitive APIsability to work autonomouslystrong collaboration with cross-functional teamsclear technical communicationPython extensions and FFIC++ developmentAPI design for complex systems
Not specified in the original listing.
Not specified in the original listing.
