Reliability Engineer, Supercomputing

Thinking Machines Lab•San Francisco, CA

5d•Onsite

About The Position

Thinking Machines Lab's mission is to empower humanity through advancing collaborative general intelligence. We're building a future where everyone has access to the knowledge and tools to make AI work for their unique needs and goals. We are scientists, engineers, and builders who’ve created some of the most widely used AI products, including ChatGPT and Character.ai, open-weights models like Mistral, as well as popular open source projects like PyTorch, OpenAI Gym, Fairseq, and Segment Anything. About the Role We're hiring an engineer to ensure the reliability of our GPU supercomputing fleet, owning the seam between hardware, firmware, and operating system. You will track the long tail of hardware issues: We are conducting frontier research in AI and a single bad NIC, HBM or a kernel driver edge case can compromise an experiment. Your job is to diagnose these issues, track their root cause down to the hardware, and resolve them internally or directly with vendors so that our researchers can run at scale and with confidence. Note: This is an "evergreen role" that we keep open on an on-going basis to express interest. We receive many applications, and there may not always be an immediate role that aligns perfectly with your experience and skills. Still, we encourage you to apply. We continuously review applications and reach out to applicants as new opportunities open. You are welcome to reapply if you get more experience, but please avoid applying more than once every 6 months. You may also find that we put up postings for singular roles for separate, project or team specific needs. In those cases, you're welcome to apply directly in addition to an evergreen role.

Requirements

Bachelor’s degree or equivalent experience in computer science, engineering, or similar.
Proficiency in at least one backend language (we use Python or Rust).
Experience operating large‑scale clusters and container orchestration systems (e.g. Kubernetes or Slurm).
Comfort operating across the stack and owning projects end-to-end.
Thrive in a highly collaborative environment involving many, different cross-functional partners and subject matter experts.
A bias for action with a mindset to take initiative to work across different stacks and different teams where you spot the opportunity to make sure something ships.

Nice To Haves

Fluency with Linux systems and debugging tools.
Proven statistical rigor in analyzing reliability.
A track record of debugging a problem from application symptom to the root cause in hardware.
Comfort reading vendor errata, firmware release notes, and kernel changelogs.
Experience engaging hardware vendors directly — not just through escalation portals.
Linux kernel literacy: the scheduler, memory management, IRQ paths, and the driver model.
Out-of-band management experience: BMC / iDRAC / IPMI / Redfish.
Depth in GPU hardware health: Xid error taxonomy, NVLink, NVSwitch, fabric manager, and DCGM.
Proficiency in at least one backend language (we use Python and Rust).
Significant ownership of the hardware reliability function at scale.
Strong writing skills for vendor cases and postmortems.
An instinct for telling apart a flaky machine, a flaky workload, and a flaky test.

Responsibilities

Investigate, reproduce, and remediate issues across large GPU clusters.
Own the drivers, kernel surface, and diagnostics that span hardware, firmware, and OS.
Automate the monitoring of fleet reliability and analyze error rates to validate whether a fix or firmware change measurably reduced failures rather than shifting them around.
Drive the firmware lifecycle: tracking, qualification, staged rollout, and regression analysis.
Engage vendors directly — GPUs, server OEMs, NIC vendors, and storage vendors — to get real fixes rather than ticket numbers.
Manage RMA flows when hardware needs to come out.
Monitor and improve GPU hardware health signals and turn them into actionable reliability improvements.
Write clear postmortems and vendor cases that move issues forward.