Senior Software Engineer, Distributed AI & Triton on GPUs

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

Benefits offered by this job

AMD benefits

Job summary

AMD is seeking an experienced engineer to advance the Triton compiler and runtime for AMD Instinct GPUs. You will design native distributed execution, optimize inter-GPU communication, and work across the compiler, runtime, and ROCm layers to push scalability and performance.

Collaboration with architecture and performance teams is essential as AMD broadens its multi-GPU AI capabilities. The role focuses on delivering industry-leading distributed GPU support, contributing to open-source

Qualifications

• Experience in GPU architecture, compiler tech, and distributed GPU systems at multi-GPU scale.
• Ability to work across compiler, runtime, and hardware layers to maximize scalability.
• Background in optimizing AI/HPC workloads on multi-GPU platforms.

Responsibilities

• Design and implement native distributed communication and execution capabilities within the Triton AMDGPU backend for scalable multi-GPU execution.
• Develop compiler and runtime mechanisms for native GPU-initiated communication including collectives and remote memory access.
• Profile, debug, and optimize cross-stack performance across compiler, runtime, ROCm, and hardware.
• Collaborate with architecture, compiler, runtime, and performance teams to enable next-gen distributed GPU programming.
• Contribute to open-source Triton and ROCm distributed ecosystem.

Skills

GPU architecture

Compiler technologies

Distributed GPU systems

Multi-GPU performance

Performance tooling

Education

Bachelor’s or Master’s Degree in Computer Engineering, Computer Science, Electrical Engineering

Tools

Triton

ROCm

NCCL

NVSHMEM

MPI

Job description

AMD is seeking an experienced engineer to advance the Triton compiler and runtime for AMD Instinct GPUs. You will design native distributed execution, optimize inter-GPU communication, and work across the compiler, runtime, and ROCm layers to push scalability and performance.

Collaboration with architecture and performance teams is essential as AMD broadens its multi-GPU AI capabilities. The role focuses on delivering industry-leading distributed GPU support, contributing to open-source