Software Engineer, ML Networking
Core
Systems-level engineer building and maintaining software that interfaces between ML accelerators and high-speed networks to optimize tensor movement and distributed system performance.
Role type
Senior IC systems/networking engineer (ML infrastructure)
Builds
High-performance network infrastructure for ML accelerators, collective algorithms, and congestion control systems
Domain
Machine Learning Infrastructure / High-Performance Computing / Network Systems
Deliverable
production ML models | infrastructure
Required skills
Network protocols (TCP/IP, XDP, eBPF, io_uring, epoll), Kernel networking, User-space networking (DPDK, RDMA), Systems programming (Rust, C/C++), Memory management, Lock-free data structures, NUMA-aware programming, Distributed systems debugging, PCIe drivers, SmartNIC programming
Preferred skills
ML accelerator drivers, New network protocol design, Graph algorithms, Compression algorithms
Technologies
Rust, C/C++, DPDK, RDMA, XDP, eBPF, io_uring, epoll, PCIe, SmartNICs
Responsibilities
Write and maintain software interfacing accelerators with high-speed networks, Diagnose and resolve networking issues in distributed systems (OSI layers 2-4), Optimize congestion control algorithms for large-scale synchronous workloads, Implement new collective algorithms to improve latency, Debug kernel-level network latency spikes
Seniority
Senior, hands-on IC
