11-08, 13:25–14:10 (US/Eastern), Central Park West
The landscape of data analytics is rapidly evolving, driven by a not-so-secret trend – concentrating more computing power, more memory, and other resources into dense machines. Over the last decade, CPU memory and core count have grown exponentially, and memory and computing capabilities of GPUs have similarly increased dramatically. Collocating many of these devices into single instances have ushered in not only optimizations but also new computing paradigms.
Most importantly, Cloud Service Providers have made these achievements widely accessible and on-demand. Easy access to powerful cloud instances means that every Python developer has the opportunity to accelerate analyses, simplify workflows, and tackle more complex problems. The densification of CPUs, GPUs, and accelerated networking is a shift change for academics and engineers who must develop techniques and change the way they write software to best leverage the next wave of computing machines.
This talk will focus on the new and evolving computing architectures and how they benefit Python and PyData. We'll reflect on the history of scientific/business computing, investigate how software (and especially Python) have co-evolved with these recent trends, and explore the potential implications of these newer architectures.
No previous knowledge expected
