Member of Technical Staff - GPU Infrastructure

Design, deploy, optimize, and support large-scale GPU infrastructure for customers, including GPU clusters, orchestration, high-performance networking, parallel filesystems, system performance, infrastructure troubleshooting, documentation, and operational support.

Responsibilities

  • Partner with clients to understand workload requirements and design GPU cluster architectures
  • Create technical proposals and capacity plans for clusters ranging from 100 to 10,000+ GPUs
  • Develop deployment strategies for LLM training, inference, and HPC workloads
  • Present architectural recommendations to technical and executive stakeholders
  • Deploy and configure SLURM and Kubernetes
  • Implement InfiniBand, RoCE, and NVLink networking
  • Optimize GPU utilization, memory management, and inter-node communication
  • Configure Lustre, BeeGFS, and GPFS filesystems
  • Tune kernel and CUDA configurations
  • Resolve customer infrastructure issues
  • Implement monitoring, alerting, and automated remediation
  • Provide 24/7 on-call support for critical customer deployments
  • Create runbooks and documentation

Requirements

  • 3+ years of hands-on experience with GPU clusters and HPC environments
  • Deep expertise with SLURM and Kubernetes in production GPU settings
  • Experience with InfiniBand configuration and troubleshooting
  • Strong understanding of NVIDIA GPU architecture, CUDA, and drivers
  • Experience with Ansible and Terraform
  • Proficiency in Python, Bash, and systems programming
  • Customer-facing technical leadership experience
  • Experience with NVIDIA drivers, Fabric Manager, and DCGM
  • Experience configuring Docker, Containerd, and Enroot for GPUs
  • Linux kernel tuning and performance optimization
  • AI workload network topology design
  • Knowledge of power and cooling requirements for high-density GPU deployments

Benefits

  • Equity incentives

See also

要針對這個職缺調整履歷嗎?

目前無法檢查您與這個職缺的符合程度;請先將履歷加入個人檔案,下次即可查看。

A new version of freehire is available