Senior Deep Learning Performance Architect - LPU
Core
Designing novel GPU and system architectures to advance AI inference performance and efficiency, and building power/performance models to guide next-gen hardware.
Role type
Senior IC deep learning performance architect (hardware-software co-design)
Builds
Next-generation GPU architectures and AI inference stacks
Domain
Semiconductor hardware + AI inference
Deliverable
production ML models
Required skills
GPU computing (CUDA), HPC (MPI, OpenMP), C/C++/Python, computer architecture, deep learning algorithms, systems-level performance modeling
Preferred skills
Systems-level profiling, publishing performance studies, developing CUDA kernels/compilers for custom ASICs
Technologies
CUDA, MPI, OpenMP, C, C++, Python
Responsibilities
Design GPU/system architectures for AI inference, construct and test deep learning algorithms, analyze hardware-software relationships, build power/performance models, collaborate with software/research/product teams
Seniority
Senior, hands-on IC
