Deep Learning Framework Communications Engineer - New College Grad 2026
Core
Integrate communication libraries (NCCL, NVSHMEM) into AI frameworks (PyTorch, TRT-LLM, vLLM) to optimize multi-GPU performance for training and inference.
Role type
Deep Learning Framework Communications Engineer
Builds
AI toolkits and communication libraries for multi-GPU clusters
Domain
Artificial Intelligence / High Performance Computing
Deliverable
production ML models
Required skills
Deep Learning Frameworks (PyTorch, JAX, TRT-LLM, vLLM), C++, CUDA, Python, AI compilers, multi-GPU communication, performance profiling, fault-tolerant system design
Preferred skills
Parallel programming (NCCL, NVSHMEM, MPI), kernel authoring (Triton, cuTe), AI compiler pattern matching, memory hierarchy understanding
Technologies
PyTorch, TRT-LLM, vLLM, SGLang, NCCL, NVSHMEM, CUDA, Triton, cuTe, Python, C++, JAX
Responsibilities
Integrate communication library features from PoC to production, analyze AI workloads for multi-GPU requirements, improve AI compilers for communication hiding, characterize performance on multi-GPU clusters, design fault-tolerant solutions, author custom kernels, influence library roadmaps
Seniority
Entry-level (New College Grad 2026)