AI & Systems Co-Design Scientist — GenAI Hardware

Meta; Menlo Park (CA); On-site; USD 122,000 - 181,000; Full time

Job summary

Meta is seeking a Research Scientist in AI & Systems Co-Design (PhD) to advance GenAI efficiency from model to hardware at datacenter scale. You will collaborate across architecture, compiler, and runtime teams to co-design deployment techniques and optimized ML kernels for accelerators.

The role requires a PhD (or equivalent) with strong publication track record, deep Python proficiency, and experience with AI frameworks, CNNs/Transformers/LLMs.

Qualifications

• Bachelor's degree completed before joining Meta
• PhD in CS/EE or related field, or Master's with 3+ years of experience
• Proven research in hardware-aware model enablement or AI system performance modeling
• Hands-on AI hardware architecture experience and on-device mapping development
• Theoretical and practical knowledge of AI models (CNNs, Transformers, LLMs)
• Experience in system-level performance analysis and benchmarking
• Proficient in Python and at least one major AI framework
• Track record of publishing research and communicating results to stakeholders

Responsibilities

• Explore and optimize parallelisms and compute efficiency for GenAI systems
• Co-design model deployment techniques for scalable hardware efficiency
• Benchmark and model performance across AI workloads for future hardware and runtime feedback
• Prototype and productionize optimized ML kernels for accelerators
• Influence the hardware roadmap for Meta's AI accelerators
• Lead cross-functional initiatives across engineering groups
• Guide AI hardware requirements and software stack design for future workloads

Skills

Python

AI framework

Research

System optimization

Education

Bachelor's degree in CS/Engineering

PhD in CS or EE

Master's with 3+ years experience

Tools

CUDA

HIP

Triton

HDL

Job description

Meta is seeking a Research Scientist in AI & Systems Co-Design (PhD) to advance GenAI efficiency from model to hardware at datacenter scale. You will collaborate across architecture, compiler, and runtime teams to co-design deployment techniques and optimized ML kernels for accelerators.

The role requires a PhD (or equivalent) with strong publication track record, deep Python proficiency, and experience with AI frameworks, CNNs/Transformers/LLMs.