Software Engineer, Correctness & Reliability, Dojo
What to ExpectAs a member of the Dojo correctness and reliability team, you will be responsible for building program analysis tools (fuzzers, compiler instrumentation passes, run-time monitoring components) for our in-house supercomputer. Join a small team of experienced developers in automating software and hardware testing.
What You’ll Do- Take ownership of tooling software for the compiler and hardware monitoring
- Develop algorithms to improve sensitivity and performance of the analysis tools
- Debug functional issues on massively parallel systems, including compiler bugs, defective chips, and hardware design bugs
- Collaborate with Dojo HW and SW teams to understand the current architecture and propose future improvements, especially where it helps reliability
- Work with Autopilot SW team to assure smooth transition of training from GPU to Dojo
- Degree in Engineering, Computer Science, or equivalent in experience and evidence of exceptional ability
- Prior industry or research experience in compilers, program analysis, and/or hardware testing
- Strong C++ and basic assembly skills