We make the best chips physically possible for the large model needs of frontier labs.
About the role
Design-quality and verification tooling: Build and maintain the linters, static checks, and schema-driven records that check our engineering rules automatically on every change. Signoff evidence and reporting: Turn CI, regression, and verification results into dashboards and reports that show how close each block is to signoff, without anyone assembling them by hand. Found on 1752vc Careers, the job board for startup and VC roles.
What they're looking for
- A strong software engineer: You write production Python and are comfortable in a second language such as Rust. You have built tools that other engineers depend on
- CI and build systems: You have worked with GitHub Actions or a similar CI system, and with Bazel or another build system used in a large repository
- Chip development experience: You have worked on or alongside a silicon team through at least one tapeout, as a design, DV, CAD, or methodology engineer. You understand how lint, clock domain crossing checks, coverage, waivers, and ECOs fit into signoff
- Judgment about process: You can tell a rule that catches problems from a rule that only slows people down, and you would rather remove the second than enforce it
- Clear writing: You can explain a rule and the reason for it in a few plain sentences
- Occasional travel: Expect a few trips a year to meet the teams you work with in person
More about this role
MatX's mission is to make the world's best AI models run as efficiently as allowed by physics, bringing the world years ahead in AI quality and availability. We build the full stack, from silicon and rack design through compilers, kernels, and ML models.
We are building a single engineering methodology for how our chips are specified, designed, verified, and signed off, together with the automation that makes it easy to follow. We are seeking Methodology Engineers to build and run that automation. You will be working in an existing system, not starting one, and the job is to connect strong team procedures and make them cheap to follow. Success means that engineers spend less time on process over time, not more.
Design-quality and verification tooling: Build and maintain the linters, static checks, and schema-driven records that check our engineering rules automatically on every change.
Signoff evidence and reporting: Turn CI, regression, and verification results into dashboards and reports that show how close each block is to signoff, without anyone assembling them by hand.
Exceptions and waivers: Build the tooling that keeps exceptions to our checks owned, justified, and current...
Browse similar: Startup jobs · Remote jobs · San Francisco Bay Area