Artificial Intelligence Infrastructures

Recommended Prerequisites

Proeficient in Programming and Fluent in English B2 Level (ideally C1) as per Common European Framework of Reference for Language skills.

Teaching Methods

Lecture classes (T): presentation and discussion around the topics of the course.

Lab classes (PL): application of theoretical concepts in practical assingments and discussion of research work.

The practical assignments may also be developed by the students in remote and asynchronous mode with remote support of teacher.

Learning Outcomes

This curricular unit aims to acquire knowledge on the management of information services and infrastructures of high performance to support the massive processing of data in Artificial Intelligent applications. It also aims to acquire knowledge on the edge computing model to support distributed learning techniques.

It is expected that the students acquire/develop the following core competencies:

•Skills on planning and management of infrastructures to support the massive data processing in applications of Artificial Intelligence.

•Skills on computational models that include hardware with low-power (GPU) and System-On-Chip (Soc) in the edge.

•Practical application of the theoretical knowledge on the planning and infrastructure management using centralized computing, as well as edge computing to support distributed Artificial Intelligence applications.

and the following secondary competencies:

•Problem solving, oral and written communication, interpersonal relations, and team work.

Work Placement(s)

No

Syllabus

1. Support infrastructures for Artificial Intelligence: an introduction

2. Managing centralized/data center infrastructures: computing, storage and communications

3. Cloud Container orchestration systems (ex. Kubernetes, Docker, Vagrant, Mesos) and Edge orchestration platforms (ex. KubeEdge)

4. Real-time big data architectures: Kappa and Lambda

5. Scalable and distributed transport (ex. Apache Kafka)

6. Big data processing frameworks (ex. Apache Hadoop and Spark)

7. Platforms for edge computing in hardware of software of last generation.

8. Edge computing on GPUs and System-on-Chips (SoCs): Introduction

Assessment Methods

Assessment
Research work: 20.0%
Exam: 40.0%
Laboratory work or Field work: 40.0%

Bibliography

• Articles, and resources available in the Internet for specific topics.

•Matei Zaharia, Patrick Wendell, Andy Konwinski, Holden Karau, “Learning Spark: Lightning-Fast Big Data Analysis”, 2nd edition, O’Reilly, July 2020.

•Justin Neroda, Steve Escaravage, Aaron Peters, “Enterprise AIOps”, O’Reilly, August 2021.

•Neha Narkhede, Gwen Shapira, and Todd Palino, “Apache Kafka: the definitive guide”, O’Reilly, 2017.

•Jan Kunigk, Ian Buss (Author), Paul Wilkinson, Lars George, “Architecting Modern Data Platforms: A Guide to Enterprise Hadoop at Scale”, O’Reilly, 2019.

•Buyya, Rajkumar, and Satish Narayana Srirama, eds. Fog and edge computing: principles and paradigms. John Wiley & Sons, 2019.

•James Urquhart, “Flow Architectures”, O’Reilly, 2021.

•Thomas Sterling, Matthew Anderson, and Maciej Brodowicz, “High Performance Computing: Modern Systems and Practices”, 1st Edition, Morgan Kaufmann, December 2017.