Software Engineer - Host networking
Description
This Software Engineer will be working on NICs and Transport solutions addressing growing demands of the distributed fleet of accelerators for our AI workloads. Do you want to work on transport for large scale AI clusters? Do you want to develop innovative solutions to our challenges and ship them into production? This role on our host networking teams is for you!
Responsibilities
Design and implement drivers (and/or Firmware) for (network) ethernet adapter functions, Transport stack for RDMA, control functions with the host/accelerators Collaborate with ASIC and HW teams, and external partners in building infrastructure scale embedded solutions Work with cross functional teams through releasing software to production and supporting them Mentor team members who will also work on building driver and firmware software
Qualifications
Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience 7+ years of experience in C/C++ 7+ years of experience in Systems programming, TCP/IP, RDMA/RoCE drivers and/or Firmware Experience with HW bringup Successful candidate must remain in role in the same team in India for a minimum period of 24 months before being eligible for transfer to another role, team or location Experience with Congestion control for RDMA/RoCE networks Experience with Qemu, FPGA Emulation environment is a plus Working knowledge of transport stack particularly RDMA (RoCEv2) on Linux Experience with HW offloads for Transport stack Experience with Linux Kernel, especially drivers and network stack Working knowledge of Collectives (XCCL) and GPU direct for AI workloads