Senior GPU Infra Architect with Equity

Sciforium; San Francisco (CA); On-site; USD 150,000 - 220,000; Full time

Benefits offered by this job

Medical, dental, and vision

401k plan

Lunch and beverages

Flexible time off

Salary and equity

Job summary

Sciforium in San Francisco is seeking a Senior HPC & GPU Infrastructure Engineer to own the health, reliability, and performance of our GPU compute cluster. You will be the primary custodian of a high-density accelerator environment, bridging hardware operations, distributed systems, and ML workflows.

This role involves hands-on Linux systems engineering, GPU driver bring-up, and maintaining the ML software stack (CUDA/ROCm, PyTorch, JAX, vLLM).

Qualifications

• 5+ years in HPC, GPU cluster ops, or related Linux roles.
• BS/MS in CS/CE/EE or related field.
• Deep knowledge of NVIDIA or AMD GPUs, drivers and kernel debugging.

Responsibilities

• Ensure system health, reliability, and performance of the GPU compute cluster.
• Lead on-call responses for outages, GPU failures, and node crashes.
• Develop monitoring for GPU health, memory errors, and topology issues.
• Coordinate vendor and data-center activities for repairs and RMAs.
• Manage Linux OS, patching, kernel tuning, and automation for large fleets.
• Secure infrastructure with VPNs, firewalls, SSH hardening, and access controls.
• Deploy and bring up new GPU nodes, BIOS, NUMA tuning, and topology validation.
• Maintain ML stacks: PyTorch, JAX, CUDA toolkit, cuDNN, ROCm, NCCL.

Skills

Bash scripting

Python scripting

Linux internals

GPU driver debugging

CUDA toolkit

ROCm stack

Networking security

RDMA networking

Education

Bachelor's or Master's degree in Computer Science/Computer Engineering/Electrical Engineering

Tools

Slurm

Kubernetes

Run:AI

Ansible

SaltStack

Terraform

NVIDIA drivers

ROCm tooling

Job description

Sciforium in San Francisco is seeking a Senior HPC & GPU Infrastructure Engineer to own the health, reliability, and performance of our GPU compute cluster. You will be the primary custodian of a high-density accelerator environment, bridging hardware operations, distributed systems, and ML workflows.

This role involves hands-on Linux systems engineering, GPU driver bring-up, and maintaining the ML software stack (CUDA/ROCm, PyTorch, JAX, vLLM).