Design, develop, and optimize compilers to meet performance goals for deep learning accelerators.
Collaborate with a global team to design, develop, optimize, and benchmark compiler performance.
Develop machine-independent and machine-specific code generation and optimization techniques for compilers.
Profile various machine learning workloads to identify and implement optimization opportunities in compilers.
Innovate and develop tools to support performance analysis and tuning, with a particular focus on compiler efficiency.
Qualifications
Bachelor's, Master's in Computer Science or related fields.
Experience of 2 to 7 years.
Experience with MLIR and LLVM compiler stacks Strong expertise with production compiler stacks and experience with code generation and optimization techniques.
Solid understanding of compiler toolchains and validation techniques.
Knowledge of loop nest transformation frameworks is highly preferred.
Familiarity with deep learning frameworks is a plus. Proficiency in C++ programming; familiarity with Python is preferred.
Excellent communication and presentation skills, with experience working in a cross-geo environment.