Internet and social media have completely changed the way in which we live and interact with other people around the world. Millions of connections and interactions happen every day at incredible speed, and graphs are a key tool to understand and describe these phenomena. The goal of the High-Performance Graph Analytics course is to introduce the students to the world of graphs and to teach how to analyze graphs at scale both in theory and in practice.The course is realized in collaboration with Oracle Labs and will be carried out with frontal lessons which introduce theoretical concepts and algorithms used in graph analytics, and with a multi-track design contest in which students will apply what they learnt to real-life applications and open problems. Students will be followed during the contest with ad-hoc tutoring sessions.In the contest, students will implement high-performance graph analytics algorithms and perform machine learning classification tasks with state-of-the-art graph embeddings algorithms.Computing horsepower and software tools (e.g. VMs, graph analytics libraries, GPUs) for the design contest will be provided by the NECST Laborsatory.Up to the best five participants to the design contest will take part to a second selection phase held directly by Oracle Labs, and have the opportunity to be selected for a one-day per week internship with Oracle from Milan, or for full-time internships at Oracle Labs in Zurich or Redwood City.

dal March 2019 a March 2019