Senior AI Performance Engineer: Scale ML at Data Center

Graphcore; Austin, Northern (TX, KY); Hybrid; USD 200,000 - 260,000; Full time

Benefits offered by this job

Medical, dental, and vision coverage

Flexible working hours

Hybrid working arrangements

Professional-development resources

Job summary

Graphcore in Austin, Texas seeks a Senior AI Performance Engineer to analyze and optimize AI training and inference workloads across large-scale distributed systems. You will own investigations from problem definition through validation, turning profiling, benchmarking and modeling results into actionable improvements for engineers.

You will design benchmarks, diagnose bottlenecks and optimize distributed communication software using MPI, NCCL, UCX and RDMA.

Qualifications

• Strong experience profiling and optimizing AI, machine learning or high-performance computing workloads.
• Experience with distributed systems and communication libraries such as MPI, NCCL, UCX or libfabric.
• Strong C++ and Python skills, including reliable tools or performance-sensitive software.
• Deep understanding of compute, memory and communication behavior in large-scale systems.
• Ability to own complex technical work and coordinate improvements across teams.
• Familiarity with MLPerf, accelerated architectures, ML frameworks or high-performance interconnects.

Responsibilities

• Own complex performance investigations from problem definition through validation.
• Design benchmarks, investigate system bottlenecks and optimize distributed communication software across technologies such as MPI, NCCL, UCX and RDMA.
• Turn profiling, benchmarking and modeling results into improvements engineers can build from.

Skills

Profiling & optimization

Distributed systems

MPI/NCCL/UCX

C++

Python

Cross-team coordination

MLPerf

Interconnects

Tools

MPI

NCCL

UCX

libfabric

Job description

Graphcore in Austin, Texas seeks a Senior AI Performance Engineer to analyze and optimize AI training and inference workloads across large-scale distributed systems. You will own investigations from problem definition through validation, turning profiling, benchmarking and modeling results into actionable improvements for engineers.

You will design benchmarks, diagnose bottlenecks and optimize distributed communication software using MPI, NCCL, UCX and RDMA.