We are seeking a motivated Partner Enablement Engineer to guide our key partners and customers with NCCL. You will engage with our partners and customers to root cause functional and performance issues reported with NCCL. Conduct performance characterization and analysis of NCCL and DL applications on groundbreaking GPU clusters. Develop tools and automation to isolate issues on new systems and platforms, including cloud platforms (Azure, AWS, GCP, etc.). Guide our customers and support teams on HPC knowledge and standard methodologies for running applications on multi-node clusters. Document and conduct trainings/webinars for NCCL. Engage with internal teams in different time zones on networking, GPUs, storage, infrastructure and support.
To be successful in this role, you will have a B.S./M.S. degree in CS/CE or equivalent experience with 5+ years of relevant experience. Experience with parallel programming and at least one communication runtime (MPI, NCCL, UCX, NVSHMEM) is required. Excellent C/C++ programming skills, including debugging, profiling, code optimization, performance analysis, and test design are necessary. Experience working with engineering or academic research community supporting HPC or AI is also required. Practical experience with high performance networking: Infiniband/RoCE/Ethernet networks, RDMA, topologies, congestion control is essential. Expert in Linux fundamentals and a scripting language, preferably Python is required. Familiarity with containers, cloud provisioning and scheduling tools (Docker, Docker Swarm, Kubernetes, SLURM, Ansible) is also required. Adaptability and passion to learn new areas and tools is necessary. Flexibility to work and communicate effectively across different teams and timezones is also required.
XML job scraping automation by YubHub
