M
Software Engineer, SystemML - AI Networking
Meta
Menlo Park, CA
≈ 1,1 млн–1,9 млн ₽ · 12,3 тыс.–20,8 тыс. €
14 167 – 23 917 $
🏢 Офис
Middle
Полная занятость
Продуктовая
Tech
Описание вакансии
Обязанности
•Analyze distributed GPU performance on RoCE and Infiniband
•Build SW benchmarks and performance tuners
•Develop AI framework and trainer components
•Improve distributed ML reliability and performance
•Lead collective communication library development
Технологии: C#, C++, CUDA, Data-parallel, Distributed Data Parallel, Distributed data, Fully Sharded Data Parallel, GPU Architecture, HPC, Infiniband, NCCL, Parallel Computing, Pipeline Parallel, PyTorch, Python, RoCE, Tensor Parallel