Senior Software Engineer
Core
Diagnose and resolve communication-related performance bottlenecks in large-scale distributed systems, specifically focusing on InfiniBand fabrics, RDMA transport layers, and HPC/AI workloads.
Role type
Senior Software Engineer (HPC/AI Infrastructure)
Builds
Optimized network performance and communication stacks for High-Performance Computing and AI training environments.
Domain
High-Performance Computing (HPC) / Artificial Intelligence (AI) / Network Infrastructure
Deliverable
production ML models
Required skills
InfiniBand architecture, RDMA, MPI, NCCL, C++, Python, C, Java, network performance tuning, GPU communication stack internals, collective algorithms (AllReduce), job schedulers
Preferred skills
NVIDIA hardware validation, custom tooling development, QoS policy tuning, adaptive routing configuration
Technologies
InfiniBand, NVIDIA, MPI, NCCL, C++, Python, C, Java, NVLink, SHARP
Responsibilities
Analyze InfiniBand fabric behavior (congestion, credit starvation, route imbalance), tune RDMA transport layers, develop custom tooling for fabric telemetry, collaborate with operations on fabric features, support benchmarking of HPC/AI workloads.
Seniority
Senior, hands-on IC
