Description
NVIDIA is hiring a Software Engineer to design, implement, and optimize scalable high-performance dense linear algebra kernels for GPU-accelerated libraries such as cuSOLVER. The role involves developing algorithms including triangular factorizations, eigenvalue and singular value decompositions; leading library engineers, QA engineers, and interns; collaborating with product management and external partners; and improving library quality, performance, and maintainability. Candidates need a PhD or MSc in a relevant field, at least five years of experience developing high-performance numerical linear algebra software with C++ and parallel programming, and experience with dense linear algebra libraries. The posting is marked hybrid and lists base salary ranges for Level 4 and Level 5 engineers.
