Senior Compute Engineer

Torus•Salt Lake City, UT

1d•$150,000 - $190,000•Hybrid

About The Position

The Senior Compute Engineer will own the design, provisioning, and lifecycle of the compute infrastructure that powers Torus Systems. This hands-on technical role takes GPU and CPU servers from rack installation to production readiness, creating repeatable, secure, automated processes for bare-metal bring-up, out-of-band management, and fleet operations at scale. The engineer sets standards for commissioning, configuring, and maintaining compute and GPU fabric across multiple sites, and mentors junior engineers as the team grows. The role works across hardware, firmware, Linux, networking, automation, security, facilities, and observability. It includes GPU node and fabric validation, power and cooling coordination, hardware procurement and evaluation, commissioning and acceptance testing, fleet automation and patching, compliance support, incident escalation, and documentation. This hybrid remote position is based in Salt Lake City, with travel to Torus locations in Salt Lake and Utah counties as needed and occasional travel to remote sites. Applicants must already be authorized to work in the United States without employer sponsorship. Candidates are subject to a background check.

Requirements

Bachelor's degree in computer science, engineering, or a related field, or equivalent practical experience.
At least 5 years of experience in systems engineering, data center operations, or infrastructure engineering, with significant bare-metal server experience.
Hands-on experience configuring and managing BMCs at scale, such as Supermicro IPMI, iDRAC, or iLO, including Redfish and IPMI tooling.
Hands-on experience with GPU or accelerated compute nodes, including drivers, firmware, and burn-in validation.
Working knowledge of high-speed server networking (25/100/400G, RDMA, RoCE, or InfiniBand), including optics, breakouts, and cabling.
Strong Linux administration skills, including networking, storage, and command-line troubleshooting.
Familiarity with automated provisioning tools and methods such as PXE/iPXE, cloud-init, MAAS, Foreman, or Ansible.
Solid understanding of server hardware, including CPUs, memory, RAID/storage controllers, NICs, and firmware.
Working knowledge of compute networking fundamentals, including VLANs, LACP/bonding, and out-of-band management networks.
Proficiency in at least one scripting language, such as Python or Bash.
Experience writing technical documentation, runbooks, and commissioning or acceptance procedures.
Ability to work independently, prioritize competing tasks, and mentor less experienced teammates.
Willingness to travel within Utah, with occasional travel to remote sites.
Willingness to participate in an on-call rotation.
Applicants must already have legal authorization to work in the United States without employer sponsorship.

Nice To Haves

Experience with InfiniBand fabric management (UFM or similar) and rail-optimized GPU cluster design.
Familiarity with liquid-cooled systems, such as CDU or direct-to-chip cooling, or high-density power distribution using switched or metered PDUs.
Experience with infrastructure-as-code and configuration management at fleet scale.
Experience with endpoint security and device management tools such as EDR or MDM.
Exposure to hardware security practices such as secure boot, TPM, firmware signing, and supply chain validation.
Experience operating bare-metal infrastructure for third-party tenants or hosting partners.
Familiarity with OT/ICS or energy infrastructure environments.
Knowledge of NIST, CIS, IEC 62443, NERC CIP, CMMC, or similar security frameworks.
Relevant certifications such as RHCSA/RHCE, Linux Foundation, or vendor server certifications.
Experience with monitoring and observability stacks such as Datadog, Prometheus, or Grafana.

Responsibilities

Configure and standardize BMCs, including network settings, user and role management, certificates, alerting, and firmware.
Define hardware-side provisioning standards for BIOS/UEFI, boot order, option ROM settings, storage, and RAID configuration, and hand off to the MAAS-based provisioning platform.
Secure the out-of-band management plane through dedicated management networks, credential rotation, disabling unused services, and access controls.
Own firmware lifecycle management for BIOS, BMC, CPLD, NICs, storage controllers, and GPUs, including testing and staged rollout.
Bring up and validate GPU nodes, including drivers and CUDA, GPU and NVLink health, diagnostics, and burn-in.
Own node-side GPU fabric configuration and validation, including NIC firmware, links, cabling and port mapping, breakouts, and fabric manager integration.
Validate optics, cables, and MPO polarity against the design before production deployment.
Work with Network Engineering on switch-side fabric configuration and cluster-level performance validation.
Own server-to-PDU outlet mapping and labeling; define staged power-on procedures that respect branch circuit limits and inrush.
Define power capping and load-balancing practices across PDU branches, and review electrical load before large power events.
Understand thermal and liquid-cooling dependencies of commissioned nodes and coordinate maintenance windows with facilities.
Establish and enforce naming, asset-tagging, and inventory standards across sites.
Define reference builds and configuration baselines for standard and GPU/accelerated compute nodes.
Evaluate and validate server hardware and work with procurement on specifications, vendor comparisons, quote reviews, and optics and cabling bills of materials.
Develop and maintain commissioning checklists and acceptance tests, including burn-in, stress testing, and hardware validation.
Verify network, storage, and compute readiness before nodes enter production.
Lead root cause analysis for failed hardware and recurring provisioning issues, and coordinate RMAs with vendors.
Define node onboarding and offboarding procedures for tenants or hosting partners, including BMC credential rotation, firmware verification, and out-of-band-only verification where a third party owns the host OS.
Build and maintain automation for hardware configuration, configuration drift detection, and fleet registration.
Integrate nodes with monitoring, logging, and endpoint security tooling, including BMC sensor telemetry and detection of wedged or flapping BMCs.
Maintain patching and update processes with minimal downtime.
Apply hardening standards such as CIS benchmarks to hosts and management interfaces.
Support NIST- and CIS-aligned audit and compliance efforts through configuration evidence and documentation.
Partner with security engineering on vulnerability remediation, including firmware-level vulnerabilities.
Collaborate with network engineering on VLANs, management network design, and site turn-up.
Collaborate with systems engineering on MAAS enlistment and commissioning, images, and telemetry integration.
Mentor junior engineers and technicians, and review their commissioning work.
Produce runbooks, diagrams, and technical documentation.
Provide escalation support for complex hardware and provisioning incidents.
Travel within Utah and occasionally to remote sites, and participate in an on-call rotation.

Benefits

Employee Rewards Package including equity.
401(k) Retirement Savings Plan.
Torus-paid life and AD&D insurance, with options to purchase additional coverage.
Voluntary short- and long-term disability insurance.
Medical plan options include traditional PPO or HSA-eligible plans, with Torus covering 50% of medical premiums.
Dental insurance.
Vision insurance.
Telehealth services, mental health support and coaching, and discounts on brands, dining, entertainment, and more for plan participants.
Unlimited discretionary PTO.
10 paid company holidays.
Waiting period-free 100% paid parental leave.
Regular company-wide offsites.
Torus-branded gear.
Peer Recognition Program.
Fully stocked break rooms with free snacks and drinks.