Senior HPC Performance Engineer: Multi-GPU Networking

NVIDIA; California (MO); On-site; USD 184,000 - 288,000; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA in the United States is seeking a Performance Engineer to advance the state-of-the-art in HPC libraries and multi-GPU communication. You will drive performance analysis across large clusters and study HW/SW interactions to optimize overall system efficiency.

You will characterize workloads, develop micro-benchmarks, and build tooling to visualize data, collaborating with teams across multiple time zones to push the cutting edge of GPU communications.

Qualifications

• MS or PhD in Computer Science or related field with perf/HPC experience.
• 3+ years of experience with parallel programming and at least one communication runtime (MPI, NCCL, UCX, NVSHMEM).
• Experience benchmarking and triage on large-scale HPC clusters.
• Good understanding of computer architecture and systems software fundamentals.
• Implement micro-benchmarks in C/C++, read/modify code when required.
• Proficient in Python scripting; able to debug across HW/SW stack.
• Familiar with containers and orchestration: Kubernetes, SLURM, Ansible, Docker.
• Adaptability and ability to collaborate across time zones.

Responsibilities

• Conduct in-depth performance characterization on large multi-GPU and multi-node clusters.
• Study interaction of libraries with HW and SW components in the stack.
• Evaluate proofs-of-concepts and trade-offs when multiple solutions exist.
• Triage and root-cause performance issues reported by customers.
• Collect performance data and build tools to visualize and analyze it.
• Collaborate with a dynamic team across multiple time zones.

Skills

Parallel programming

Performance benchmarking

Systems software fundamentals

C/C++ micro-benchmarks

Python scripting

Container orchestration

HW/SW stack debugging

Education

MS or PhD in Computer Science

Tools

Kubernetes

SLURM

Ansible

Docker

Job description

NVIDIA in the United States is seeking a Performance Engineer to advance the state-of-the-art in HPC libraries and multi-GPU communication. You will drive performance analysis across large clusters and study HW/SW interactions to optimize overall system efficiency.

You will characterize workloads, develop micro-benchmarks, and build tooling to visualize data, collaborating with teams across multiple time zones to push the cutting edge of GPU communications.