Large Scale Distributed AI

This presentation explores JVM capabilities on ARM and X86 architectures, as well as large scale data mining on embedded GPUs such as Jetson Xavier and TX2. I will also present a benchmarking demo comparing CPUs vs. GPUs for processing time-series data, using a combination of Java and Python libraries. Experience the Power of Java and the joy of Machine Learining with a cluster 125 Raspberry Pis vs. 512-Core embedded GPU- Given the resource constraints and limited connectivity of IoT devices, so far, we have only managed to integrate a small set of pre-defined functions into these objects. That is only a fraction of Machine learning techniques, which is a small subset of AI. Deep learning, on the other hand, is the backbone of many AI technologies that enables machines to mimic human perception of events and their environment through Artificial Neural Networks. This talk will present the deep learning capabilities of IoT devices as a cluster, by moving away from processing simple sensory data to deriving meaningful reasoning by forming a group of intelligent devices as a single unit using 125 Raspberry Pis vs. 500+ GPU cores.