Job closed
Ref: HP031-1821
Job description / Role
DGX Cluster Administrator
Our team is looking for a deeply technical HPC cluster administrator to manage a cluster of DGX servers. Be part of the DGX platform SW team to deliver next generation HPC and Deep Learning platforms.
What you'll be doing:
• Administer DGX servers and lab network, automate workload management, system updates and availability, actively communicate with management regarding any problems with the equipment and propose resolution plan, build and install/upgrade new systems that support company's software teams, deploy and maintain cluster management and scheduling tools.
Requirements
What we need to see:
• You have a BA, BS, or MS in CS, EE, CE or additional equivalent work experience 5 years of previous experience deploying and administrating HPC clusters,
• familiar with resource scheduling managers (Slurm (preferred), LSF, etc) demonstrated ability to script in bash and python.
• Experience with containers (Docker, Singularity, LXC) and container orchestration technologies like Kubernetes.
• Experience in designing and developing scripts/tools for continuous integration and continuous deployments for containers.
• Automating configuration management, infrastructure, and application deployments in a toolset such as Puppet, Chef, Ansible. Deep understanding of operating systems, computer networks, High Performance Applications and Deep Learning Framework (Tensorflow, Pythorch, Caffee, CNTK etc).
• Ability to work well with developers & test engineers. Passionate dedication to provide quality in support for your users.
Ways to stand out from the crowd:
• Familiarity with GPU usage in Compute Cluster and CUDA
• Experience with knowledge of Deep Learning.
• Experience with NVIDIA software for cluster management and provisioning such as nvsm, dcgm and DeepOps.
About the Company
Executive Solutions is a different breed of talent consultancy. Driven by a unique philosophy, we think differently and act differently from the pack. Were here to do one thing: to improve our clients businesses by finding, focusing and delivering the most talented people.
We dont just do traditional contingency recruitment. Discovering talent is much more than just matching a CV to a vacancy - its about understanding companys corporate culture and ensuring a comprehensive match. We believe that in order to find the right candidate we must understand the motivations and ambitions of senior management, which is reflected in our personal approach when partnering with a client.
Executive Solutions is your trusted talent solutions partner head quartered in Dubai, specialising in direct talent, tailored talent and HR talent solutions, along with Nationalisation initiatives throughout the Middle East. Driven by a unique philosophy, we incorporate innovative and comprehensive talent management solutions each tailored to suit what our clients need.