Senior HPC Applications Engineer

This is a fully remote position, open to applicants in Illinois.

📋 Description

• Take ownership of the software stack and enhance user experience on Parallel Works platforms.

• Act as the point of escalation for build failures, unsuccessful multi-node jobs, and underperforming tasks.

• Compile and package MPI implementations, compilers, and scientific libraries utilizing Spack or EasyBuild with Lmod module trees.

• Install requested AI and ML frameworks and facilitate multi-node GPU launches.

• Troubleshoot NCCL and collective behavior, mismatches between containers and drivers, storage throughput issues, and node faults.

• Conduct scaling studies and profile workloads using Nsight, VTune, TAU, HPCToolkit, or Score-P.

• Convey systems-level findings to the systems team when resolutions are necessary in the fabric or filesystem.

• Transition customer codes to new GPU architectures and venues, utilizing containers where applicable.

• Manage ticket triage, identify root causes of failed jobs, and provide written clarifications.

• Develop user guides, conduct office hours, and provide training for user communities and Government events.

• Collaborate directly with customers across on-premises clusters, cloud environments, and commercial GPU providers.

⛳️ Requirements

• A minimum of 10 years of experience supporting scientific or AI application users on Linux HPC systems.

• Proficiency in building complex software from source on both RHEL family and Debian or Ubuntu systems.

• Familiarity with compilers, MPI, CMake, autotools, and dependency management.

• Experience running Spack or EasyBuild and Lmod in a production environment.

• Knowledge of multi-node GPU workloads from a platform perspective.

• Capability to differentiate between framework issues and environment-related problems.

• Experience with site-provided software stacks on on-premises systems and cloud images.

• Working knowledge in at least one application domain: weather and climate, computational fluid dynamics, molecular and materials science, or structural analysis.

• Must be a United States citizen and eligible for a Secret clearance.

• Prior experience in a user-facing role at a Government supercomputing center, national laboratory, or university HPC center is preferred.

• Preferred experience with profiling and debugging tools, including GPU profilers, CPU profilers, gdb, and MPI tooling.

• Hands-on experience with distributed training or inference using PyTorch DDP or FSDP, DeepSpeed, Megatron-style frameworks, JAX, vLLM, or TensorRT-LLM is preferred.

• Experience with Jupyter, remote visualization, or virtual desktop support is preferred.

🏝️ Benefits

• Medical, vision, and dental coverage.

• 401(k) with company match.

• Short term disability.

• Generous paid vacation and sick time.

• Equal opportunity employer.