Software Engineer- GPU/AI/ML
Core
Building high-performance AI software stacks and optimizing GPU kernels for training, alignment, and inference of foundation models and agentic systems on AMD GPUs.
Role type
Senior IC GPU/AI/ML software engineer
Builds
ROCm ecosystem, GPU kernels, distributed training/inference pipelines for LLMs
Domain
AI/ML systems, GPU computing, high-performance computing
Deliverable
production ML models
Required skills
Modern C++, GPU architecture, kernel optimization (HIP/CUDA), distributed systems, LLM lifecycle (SFT/RLHF/GRPO), profiling tools
Preferred skills
ROCm/HIP ecosystem expertise, RTL design/Verilog, publications in AI/ML or systems optimization
Technologies
C++, HIP, CUDA, ROCm, PyTorch, TensorFlow, JAX, cuBLAS, cuDNN, CUTLASS, Thrust, CUB, NCCL, ROCm Profiler, Nsight
Responsibilities
Design and optimize GPU kernels for large-scale AI workloads; improve training and inference pipelines for foundation models; co-design hardware and software for new accelerators
Seniority
Senior, hands-on IC
