Graphcore
Principal AI Performance Engineer
Austin, Texas, United States · Principal · Full-time · Posted Oct 9, 2026
About the job
Defines performance strategy and leads analysis and optimization of AI training and inference workloads across data-center-scale systems.
What you'll do
- Lead performance investigations across models and multi-node scaling, using profiling, benchmarking, and modeling to guide engineering priorities.
- Set measurement standards and guide distributed communication strategy across teams.
- Review critical C++ and Python tools and influence decisions across hardware, software, networking, and system architecture.
What they ask for
- Deep experience profiling and optimizing AI, machine learning, or high-performance computing workloads at scale.
- Proven technical leadership across complex performance-engineering or system-architecture initiatives.
- Expert C++ and Python skills, including reliable tools or performance-sensitive software.
- Expert understanding of compute, memory, communication, and scaling behavior in distributed systems.
Nice to have
- Experience with MPI, NCCL, UCX, libfabric, or MLPerf.
- Experience with accelerated architectures or high-performance interconnects.
Tools
C++, Python, MPI, NCCL, UCX, libfabric, MLPerf
AI focus
Analyzes and optimizes AI training and inference workloads, including performance and scaling across large-scale systems.
Model-written summary · the original posting is authoritative