Senior Compiler Lead for ML Acceleration

Socket.dev; San Jose (CA); Hybrid; USD 180,000 - 240,000; Full time

Job summary

AMD is hiring a Principal Software Development Engineer to lead compiler technology and MLIR-based infrastructure for AI workloads. You will optimize model compilation, data movement, and hardware-specific code generation for NPUs and accelerators, guiding architectural direction across teams.

Role requires deep expertise in compiler design, MLIR/LLVM, and large-scale C++ codebases, with a track record of production ML infrastructure contributions.

Qualifications

• Strong experience in compiler development (front-end, middle-end, and/or back-end) and code optimization.
• Hands-on expertise with MLIR and/or LLVM-based compiler infrastructure.
• Proven experience working on neural network workloads.
• Deep understanding of graph-level and compiler-level optimizations for ML models.
• Strong C++ development skills and experience in large codebases.
• Ability to operate at Principal-level technical scope, influencing architecture and direction.
• Experience targeting or optimizing for NPUs or specialized AI accelerators.
• Familiarity with model compilation stacks and custom lowering pipelines.
• Contributions to compiler or ML infrastructure in production environments.

Responsibilities

• Lead architecture design and development of compiler components and optimization pipelines for machine learning.
• Design and implement MLIR-based compiler flows to lower high-level ML representations into optimized hardware-specific code.
• Drive model compilation and data movement optimization for ML inference workloads.
• Define and implement compiler strategies for operator fusion, memory planning, scheduling, and performance optimization.
• Work cross-functionally with hardware, runtime, frontend, and systems teams to align compiler capabilities with evolving accelerator architectures.
• Provide technical leadership and mentorship, influencing compiler direction and best practices.
• Contribute to long-term roadmap decisions for compiler and ML acceleration software.

Skills

Compiler development

MLIR/LLVM

Neural networks

C++

Graph optimizations

Model compilation

NPU/AI accelerators

Lowering pipelines

Production ML infra

Education

Bachelor’s/Master’s in CS/CE/EE

Tools

MLIR

LLVM

Job description

AMD is hiring a Principal Software Development Engineer to lead compiler technology and MLIR-based infrastructure for AI workloads. You will optimize model compilation, data movement, and hardware-specific code generation for NPUs and accelerators, guiding architectural direction across teams.

Role requires deep expertise in compiler design, MLIR/LLVM, and large-scale C++ codebases, with a track record of production ML infrastructure contributions.