HPC Containers: Challenges and Solutions

Date and Time: 
Wednesday April 10th 2019
CG Auditorium
Smahane Douyeb
In the age of rapid software development, deployment, and monitoring with scalable hardware, where modularity and parallelism play a considerable role, a new concept of software and technology has come into focus. Containers have been formalized and introduced into the ecosystem. The purpose of these containers is to foster the need for functional repeatability, modularity, portability, and most importantly, scalability. With applications moving towards a cloud infrastructure, individuals have the power of multiple nodes at their beck and call. High Performance Computing (HPC) is not immune to this motion towards the cloud. HPC traditionally uses job managers such as SLURM and LSF. This talk will focus on the impact (challenges) of containers in HPC with regards to single node, packaging, multi-node and diagnostic tools and cloud and bare metal compute environments.
Speaker Description: 
Smahane Douyeb is currently working as a Software Apps Engineer at Intel. Part of her job is to create and run containers for various HPC platforms for performance analysis and competitive testing purposes. She also works on HPC Python apps optimization on some Intel platforms. On her free time, she like to run, swim, hike and learn new skills.

Event Category: