NVIDIA is looking for a Senior System Software Engineer specializing in NCCL Partner Enablement. You will play a key role in supporting groundbreaking GPU clusters by engaging with partners to solve their most challenging technical problems. This position involves deep performance analysis, tool development, and knowledge sharing within NVIDIA's dynamic ecosystem.
What You'll Do
- Engage with partners and customers to root cause functional and performance issues reported with NCCL.
- Conduct performance characterization and analysis of NCCL and deep learning applications on new GPU clusters.
- Develop tools and automation to isolate issues on new systems and cloud platforms like Azure, AWS, and GCP.
- Guide customers and support teams on HPC knowledge and standard methodologies for running applications on multi-node clusters.
- Document and conduct trainings or webinars for NCCL technologies.
- Engage with internal teams across different time zones on networking, GPUs, storage, infrastructure, and support.
What We're Looking For
- A B.S./M.S. degree in CS/CE or equivalent experience with 5+ years of relevant work.
- Experience with parallel programming and at least one communication runtime like MPI, NCCL, UCX, or NVSHMEM.
- Excellent C/C++ programming skills, including debugging, profiling, code optimization, performance analysis, and test design.
- Experience working with an engineering or academic research community supporting HPC or AI.
- Practical experience with high performance networking: Infiniband, RoCE, Ethernet, RDMA, topologies, and congestion control.
- Expert in Linux fundamentals and a scripting language, preferably Python.
- Familiar with containers, cloud provisioning and scheduling tools like Docker, Kubernetes, SLURM, and Ansible.
- Adaptability and passion to learn new areas and tools.
- Flexibility to work and communicate effectively across different teams and time zones.
Nice to Have
- Experience conducting performance benchmarking and developing infrastructure on HPC clusters.
- Prior system administration experience, especially for large clusters.
- Experience debugging network configuration issues in large scale deployments.
- Familiarity with CUDA programming and/or GPUs.
- Good understanding of Machine Learning concepts and experience with Deep Learning Frameworks such as PyTorch or TensorFlow.
- Deep understanding of technology and passionate about what you do.
Technical Stack
- Languages: C/C++, Python
- Communication Runtimes: NCCL, MPI, UCX, NVSHMEM
- Systems: Linux, Docker, Kubernetes, SLURM, Ansible
- AI/ML: CUDA, PyTorch, TensorFlow
- Networking: Infiniband, RoCE, Ethernet, RDMA
- Cloud Platforms: Azure, AWS, GCP
Team & Environment
You will join the team that develops NVIDIA's GPU communication libraries, including NCCL, NVSHMEM, and GPUDirect technologies.
Benefits & Compensation
- Highly competitive salaries
- Extensive benefits package
NVIDIA is dedicated to pushing the boundaries of technology and promotes diversity, inclusion, and flexibility. As an equal opportunity employer, we are committed to fostering a supportive and empowering workplace for all.
