AI/HPC Performance Engineer: Scale Large AI Clusters

Meta; Menlo Park (CA); On-site; USD 154,000 - 217,000; Full time

Benefits offered by this job

Bonus

Equity

Benefits

Job summary

Meta is seeking an AI/HPC System Performance Engineer to drive end-to-end performance characterization of large-scale AI training and inference clusters. You will profile workloads, build dashboards, and resolve performance regressions across RDMA fabrics, NCCL/MPI, and network infrastructure.

The role sits on the Network Infrastructure Engineering team and collaborates with hardware and AI research groups.

Qualifications

• 5+ years coding experience in C, C++, Python or similar languages.
• Experience profiling and optimizing distributed AI or HPC workloads with GPU interconnects, RDMA, NCCL/MPI.
• Experience debugging complex performance issues across multi-layer systems.

Responsibilities

• Profile and benchmark AI training/inference workloads on large-scale HPC clusters to identify bottlenecks.
• Develop performance analysis frameworks and dashboards for system-level metrics (GPU utilization, bandwidth, latency).
• Investigate performance regressions in distributed AI/HPC environments including RDMA fabrics and scheduling.
• Collaborate with network, hardware, and AI teams to validate new HPC configurations.
• Design and execute capacity/scalability experiments to inform network topology decisions.

Skills

C/C++/Python

Performance profiling

Distributed AI workloads

Cross-functional work

Education

Bachelor's degree in CS/CE/related field

Tools

NCCL

MPI

RDMA networking

PyTorch

Job description

Meta is seeking an AI/HPC System Performance Engineer to drive end-to-end performance characterization of large-scale AI training and inference clusters. You will profile workloads, build dashboards, and resolve performance regressions across RDMA fabrics, NCCL/MPI, and network infrastructure.

The role sits on the Network Infrastructure Engineering team and collaborates with hardware and AI research groups.