Folkert Huizinga

This talk presents an overview of the open source project "Leela Chess Zero" (Lc0), a new type of chess engine based on the AlphaZero paper released by DeepMind. Lc0 learns to play chess solely by playing against itself without any human knowledge or intervention. Starting from scratch with random moves, its play improves each iteration using a powerful combination of reinforcement learning and neural networks. We rely both on an active open community for the code and computing resources as many petaflops are required to train a solid neural network. Currently the engine is among the four best in the world.

Over the last few years, radio astronomy has undergone a revolution: complex analogue machinery such as dishes is increasingly being replaced by arrays of simple antennae combined with fast networking and high-performance computing. The AARTFAAC project is taking advantage of this to build a round-the-clock monitor of the visible sky, providing a groundbreaking capability to detect to some of the most exciting and energetic astronomical events as they happen. However, turning the signals recorded by the telescope's antennae into astronomically relevant images is highly computationally intensive, and requires the development of a fast, flexible, scalable streaming data processing pipeline. I will describe the design of this pipeline, explaining how we use open source software to address some of the challenges we face and push forward our understanding of the Universe.