LIDA Data Science Infrastructures seminar

As data grows in scale and complexity across scientific disciplines, the ability to extract timely insights hinges on computational infrastructures that can keep pace.

This talk with Professor Karim Djemame explores the convergence of High-Performance Computing (HPC) and data science at the exascale frontier - where performance, scalability, and adaptability redefine what's possible. We highlight the architectural advances and software frameworks enabling massive parallelism, efficient I/O, and intelligent scheduling to support end-to-end data-intensive workflows. Case studies from climate modelling and real-time analytics illustrate how exascale platforms are accelerating discovery.

We also examine persistent challenges, including system heterogeneity, energy efficiency, and the co-design of algorithms with evolving hardware. Ultimately, this talk not outlines exascale computing power, but demonstrates its truly usability for data science at unprecedented scales.

This is a hybrid seminar and if you are joining online the link will be shared in advance.