Research Scientist, Systems ML - HW/SW Co-Design

What you'll do

• Translate production ML workload insights into accelerator architecture decisions and evaluate performance, power, and cost trade-offs.
• Build models, benchmarks, and tooling for architecture comparisons and pre-silicon decisions.
• Lead cross-team initiatives, establish evaluation methods, and mentor colleagues.

What they ask for

• 7+ years of industry experience (or equivalent).
• Experience in hardware/software co-design, AI accelerator architecture, ML systems, high-performance computing, or performance modeling; and experience with PPA trade-offs.
• Experience contributing to at least one silicon tapeout, from architectural exploration through pre-silicon validation.
• Understanding of modern ML workloads and how architecture choices affect performance at scale.

Nice to have

• PhD degree in Computer Science, Computer Engineering, or a related field.
• Published research at top venues (ISCA, MICRO, HPCA, ASPLOS, MLSys) or equivalent industry contributions.

Pay

$183,997/year to $257,000/year USD base salary; bonus and equity amounts not stated

Tools

MTIA, ML, large language models, generative AI, PyTorch, networking

AI focus

Co-design AI accelerators for production ML workloads, including large language models, recommendation systems, and generative AI; the posting also mentions AI skill development, AI tools, and responsible AI practices.