← All roles

GPU Cluster Engineer, Hardware Operations

Sciforium· Engineering
Apply on company site
San Jose, CA, San Francisco, CAFull-timePosted 5mo ago

About the role

Hardware Operations Engineer responsible for the physical health and foundational infrastructure of Sciforium's GPU clusters, covering data center coordination down to the base Linux OS layer to ensure a stable, secure, and fully operational environment for research and product workloads.

Responsibilities

  • System Health & Hardware Reliability: On-Call Response for outages, proactive cluster monitoring of GPU thermals, power draw, and system loads; Vendor Liaison with data center staff and third-party hardware vendors for RMAs, repairs, and maintenance; Hardware Deployment including rack and bring-up of new GPU nodes with integrated power and network connectivity.
  • Linux & Network Administration: OS Management (Ubuntu/CentOS/RHEL) across bare-metal servers; Security & Access through edge/internal networking, firewalls, VPNs, and SSH controls; Identity & Storage Management administering LDAP/Active Directory and ensuring NFS/GPFS/Lustre storage is mounted and permissioned.

Qualifications

  • 3+ years of Linux Systems Administration (boot processes, systemd, disk management)
  • Strong background in server hardware troubleshooting in high-density environments (power, cooling, PCIe)
  • Experience managing networking security (VPNs, iptables/firewalld, VLANs) and directory services (LDAP/FreeIPA/Active Directory)
  • Proficiency in Bash scripting for automation

Skills

Required:Linux System AdministrationHardware TroubleshootingNetworking SecurityLDAP/Directory Services

Benefits

  • Medical, dental, and vision insurance
  • 401k plan
  • Daily lunch, snacks, and beverages
  • Flexible time off
  • Competitive salary and equity
Last updated Aug 7, 2026. Interested? Apply on the company site.

About Sciforium

Byte-Native AI Models & Infrastructure

Sciforium is pioneering the future of AI infrastructure and research. We're developing byte-native multimodal foundation models while delivering serverless LLM serving at a fraction of traditional costs

Company profileWebsite

More open roles at Sciforium