HPC Systems Engineer, Consumer Products

About the Team

The Consumer Products Infrastructure team builds and operates the high-performance computing platforms that support product design, simulation, and validation across OpenAI’s consumer-facing hardware and software efforts. We partner closely with product engineers, designers, and applied scientists to ensure fast iteration cycles, predictable performance, and highly available systems that directly impact shipped products.

About the Role

We are seeking an experienced HPC Systems Engineer to design, scale, and operate large-scale HPC environments that power simulation-driven product development. These clusters support 1,000+ compute nodes per environment and are used heavily for engineering simulations such as FEA, CFD, multiphysics, and optimization that inform real-world product decisions.

In this role, you will work directly with consumer product and simulation engineers to optimize scheduling, reduce iteration latency, ensure license availability, and maintain highly reliable on-prem and hybrid-cloud HPC systems that accelerate product timelines.

In this role, you will:

• Architect, deploy, and operate large-scale HPC clusters (1,000+ nodes) supporting simulation workloads critical to consumer product development (Abaqus, Ansys, COMSOL, LS-DYNA, and similar tools).

• Optimize workload management using NC, IBM/Platform LSF, and Slurm, with a focus on throughput, fairness, and minimizing queue wait times for product teams.

• Design and implement strategies for workload balancing, cluster federation, and multi-scheduler environments that support diverse product workflows.

• Partner closely with product design, mechanical, electrical, and simulation engineers to debug jobs, improve parallel scaling, and accelerate design-to-validation cycles.

• Administer and harden Linux-based HPC systems (RHEL, Rocky Linux, AlmaLinux), including patching, kernel tuning, and performance optimization.

• Operate and optimize software licensing infrastructure (FlexLM, DSLS, LUM, RLM) to maximize utilization and prevent license-related development bottlenecks.

• Deploy and manage Azure CycleCloud and/or TotalCAE to enable elastic capacity, cloud bursting, and hybrid HPC workflows during peak product development cycles.

• Configure and tune high-speed interconnects, including InfiniBand (HDR/EDR/FDR), to support low-latency, tightly coupled simulation workloads.

• Design and maintain high-performance storage systems (NFS, DFS, Lustre, GPFS / Spectrum Scale, BeeGFS, Azure NetApp) optimized for simulation I/O patterns.

• Build automation and internal tooling using Python and Bash to streamline provisioning, monitoring, diagnostics, and job submission workflows.

• Implement monitoring, alerting, and capacity planning systems (Grafana/Prometheus, Slurm accounting, LSF monitoring) to ensure predictable performance for product teams.

• Ensure high availability and resiliency across globally distributed HPC clusters and hybrid cloud environments.

• Manage authentication, networking, and secure access (LDAP/AD integration, networking, VPNs).

• Produce clear documentation on cluster architecture, policies, and best practices tailored to consumer product engineering workflows.

Minimum Qualifications:

• 7+ years of experience designing and operating large-scale HPC clusters (1,000+ nodes).

• Deep expertise with NC, IBM/Platform LSF, and Slurm workload managers.

• Strong Linux system administration experience (RHEL-family preferred).

• Hands-on experience with MPI, parallel scaling, and performance tuning for simulation workloads.

• Experience using Azure CycleCloud to provision and manage HPC clusters in hybrid cloud environments.

• Proven experience operating InfiniBand or other high-speed interconnects.