Senior HPC Site Reliability Engineer

Job description

We are now looking for a Senior HPC Site Reliability Engineer to join our mission and continue improving our HPC infrastructure. A meaningful part of NVIDIA’s strength is our unique and advanced development tools and environments that enable our incredible pace of innovation. We are looking for architects to help us evolve the way our private compute cloud is architected and optimized. What you will be doing: Provide leadership in the design and implementation of our large-scale compute cloud that enables the world's top chip modelers, designers, and deep learning experts to invent groundbreaking technology. Identify architectural changes or completely innovative approaches in our cloud architecture and design. Help with strategic challenges we encounter, including: effective resource utilization in a heterogeneous compute environment, evolving our private/public cloud strategy, capacity modeling, and planning for multi-year growth and scaling across our global computing environment! What we need to see: B.sc in Computer Science, Electrical Engineering or related field or equivalent experience 8+ years of experience designing and operating large scale compute infrastructure. Experience with job schedulers such as IBM/Platform LSF, SGE, SLURM, Marathon, Chronos. Solid understanding of cluster configuration managements tools – Ansible, Puppet, Chef, Salt. Good experience providing compute services using a public cloud (AWS, Azure, Google Cloud) Strong script-writing skills: Python, Bash, Perl Knowledge of and/or experience deploying PaaS microservices – Docker, Docker Swarm, Kubernetes Understanding of fast distributed and network attached storage solutions and Linux file systems, ability to recommend and implement solutions to improve OS performance and reliability. Ways to stand out from the crowd: Linux certification from a well-known vendor - RedHat, Oracle, etc. Prior experience managing large-scale Kubernetes deployment in production. Strong skills in modern container networking and storage architecture.