Senior AI Systems Engineer: GPU Training & Inference

AMD; Markham; On-site; CAD 140,000 - 190,000; Full time

Job summary

AMD is seeking a Senior Software Development Engineer to own the end-to-end model execution stack on AMD Instinct GPUs, covering training infrastructure at scale and high-performance inference serving. You will ship LLMs on real hardware, optimize GPU kernels, and ensure frontier models run efficiently across generations.

Lead multi-node training, develop observability tooling, and collaborate with silicon teams to validate software stack integration for both training and inference workloads.

Qualifications

• Experience shipping AI/ML infrastructure for training and inference.
• Proven ability to work across distributed GPU clusters and HPC workflows.
• Familiarity with pre-silicon validation or hardware-software co-verification is a plus.

Responsibilities

• Enable and optimize large-scale model training on AMD Instinct GPU clusters with Kubernetes orchestration.
• Develop and maintain training infrastructure, including data loading and distributed checkpointing.
• Debug training issues and optimize RCCL communications across multi-node topologies.
• Implement end-to-end validation, monitoring, and compliance tooling for scale.

Skills

Large-scale training

GPU kernels

Kubernetes

Distributed training

Observability

Open ROCm

Education

Bachelor’s or Master’s degree in Computer Science

Tools

HIP

Triton

MLIR

DeepSpeed

Megatron-LM

Job description

AMD is seeking a Senior Software Development Engineer to own the end-to-end model execution stack on AMD Instinct GPUs, covering training infrastructure at scale and high-performance inference serving. You will ship LLMs on real hardware, optimize GPU kernels, and ensure frontier models run efficiently across generations.

Lead multi-node training, develop observability tooling, and collaborate with silicon teams to validate software stack integration for both training and inference workloads.