Staff Modeling Architect
Neurophos · Austin, Texas · Sunnyvale, California
Infrastructureonsitestaff$250k–$290k
At a glance
Highlights
- Own workload-to-silicon modeling for photonic AI inference chips at scale.
- Guide architecture decisions on tiling, memory, and multi-chip mapping.
- Mentor engineers while enabling software bring-up before tape-out.
What they offer
- 100% coverage of base health plan premiums plus HSA contributions
- Unlimited PTO
- 401(k) matching and stock option opportunities
- Dental, Vision, Life and supplemental insurance options
Ideal candidate
A staff-level modeling architect who thrives in co-design and owns complex problems end-to-end. You love turning cutting-edge AI workloads into clear architectural insight and mentoring others.
Requirements
- Develop Python energy and latency models covering operators, tiling and memory traffic
- Implement bit-accurate C++ functional models for bring-up before tape-out
- Contribute to C++ event-driven simulation kernel with coroutines and timed components
- Build cycle-approximate and cycle-accurate PPA models aligned with RTL
- Experience inside discrete-event or cycle-accurate simulators like SystemC, gem5 or SST
Lab profile
About Neurophos
Building silicon photonics chips for energy-efficient AI inference compute.
What they build
- Silicon photonics inference compute chips
Known for
- Silicon photonics
- Analog and SoC design
- Quantization and numerical optimization
- Performance modeling and benchmarking
Posted · Last seen