Senior Compiler Engineer – AI Accelerators; LLVM​/MLIR

WHAT YOU DO AT AMD CHANGES EVERYTHING

At AMD, our mission is to build great products that accelerate next‑generation computing experiences—from AI and data centers, to PCs, gaming and embedded systems. Grounded in a culture of innovation and collaboration, we believe real progress comes from bold ideas, human ingenuity and a shared passion to create something extraordinary. When you join AMD, you’ll discover the real differentiator is our culture.

We push the limits of innovation to solve the world’s most important challenges—striving for execution excellence, while being direct, humble, collaborative, and inclusive of diverse perspectives. Join us as we shape the future of AI and beyond.

We are seeking a hands‑on Compiler Engineer to help build and optimize AMD’s compiler stack for Neural Processing Units (NPUs). This role focuses on frontend and mid‑end compiler development for AI workloads, including graph transformations, operator fusion, tiling, memory‑layout‑aware optimizations, and efficient mapping to accelerator hardware. You will work with open‑source compiler infrastructure such as LLVM and MLIR and collaborate closely with hardware, runtime, and model teams to deliver high‑performance, production‑ready software for AMD AI devices.

The successful candidate will bring strong software engineering fundamentals, practical compiler experience, and a solid understanding of hardware/software interaction.

We welcome candidates with experience in architecture‑aware software optimization, including compiling for accelerators, ASICs, FPGAs, GPUs, NPUs, or other specialized hardware, as well as optimizing software based on hardware characteristics such as memory hierarchy, parallelism, and data movement.

This is not a pure machine learning application or data science role.

• Design, develop, and maintain compiler components targeting AMD AI accelerators, with an emphasis on frontend and mid‑end transformations.
• Implement and enhance compiler optimizations, including operator fusion, tiling, memory layout transformations, scheduling, and code generation support for AI workloads.
• Use C++ and Python to develop compiler passes, test infrastructure, and development tools on top of LLVM/MLIR and related open‑source technologies.
• Analyze model and benchmark performance, debug compiler issues, and identify efficient mappings of workloads to hardware.
• Collaborate with hardware architects, runtime teams, and machine learning framework teams to improve compiler quality, correctness, and performance.
• Contribute to build, test, and development infrastructure to improve robustness, scalability, and engineering efficiency.
• Take ownership of technically significant projects while collaborating effectively across distributed teams.

• Bachelor’s degree with 8+ years, Master’s degree with 6+ years, or PhD with 3+ years of experience in compiler engineering, systems software, or performance‑critical software development.
• Strong C++ programming skills and proficiency in Python.
• Understanding of compiler fundamentals, including intermediate representations, optimization passes, code generation, and debugging.
• Experience with LLVM, MLIR, TVM, or similar compiler frameworks.
• Experience implementing compiler passes within LLVM/MLIR or comparable compiler infrastructure.
• Knowledge of software optimizations related to hardware architecture characteristics, including memory hierarchy, data movement, threading, vectorization, or accelerator execution models.
• Experience with AI/ML model compilation or deployment, or with compiling for specialized hardware such as GPUs, NPUs, DSPs, ASICs, or FPGAs.
• Strong problem‑solving, debugging, and communication skills.

• Experience with ONNX, PyTorch, or other machine learning frameworks and their compiler/runtime interfaces.
• Experience with graph‑level transformations, domain‑specific intermediate representations (IRs), or ML‑focused optimizations.
• Contributions to open‑source compiler or AI infrastructure projects.
• Experience working in Linux‑based development environments and with modern…