OpenAI
Software Engineer, Inference – AMD GPU Enablement
About this role
OpenAI is seeking a Software Engineer to optimize and scale inference infrastructure on AMD GPU platforms. You'll work across the full stack—from kernel-level performance tuning to distributed execution—to enable efficient deployment of large AI models on AMD accelerators.
What you'll do
- Own bring-up, correctness, and performance optimization of OpenAI's inference stack on AMD hardware
- Integrate model-serving infrastructure like vLLM and Triton into GPU-backed systems
- Debug and optimize distributed inference workloads across memory, network, and compute layers
- Develop and optimize high-performance GPU kernels using HIP, Triton, or similar frameworks
- Build and tune collective communication libraries (e.g., RCCL) for multi-GPU parallelization
- Validate correctness, performance, and scalability on large GPU clusters
What they're looking for
- GPU kernel development (HIP, CUDA, or Triton)
- Distributed inference systems and multi-GPU scaling
- Communication libraries (NCCL/RCCL)
- Low-level GPU performance optimization
- End-to-end performance debugging across hardware and software layers
- GPU profiling tools (Nsight, rocprof, perf)
- Model/tensor parallelism and mixed precision
- C++ or Python systems programming
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
OpenAI
OpenAI builds AI infrastructure and products, including large-scale data center campuses for AI computing and generative AI applications for enterprise customers. The company is hiring civil engineers, project engineers, electrical design engineers, data center R&D engineers, and AI deployment engineers to expand its infrastructure capabilities and help customers deploy AI solutions.
View all jobs at OpenAILikely interview questions
- Tell us about your experience writing or porting GPU kernels using HIP, CUDA, or Triton. What was the most challenging performance optimization you've tackled?
- Describe your experience with distributed inference systems. How have you approached debugging performance bottlenecks across memory, network, and compute layers?