More efficient big data infrastructures
Big data requires high-performance infrastructure, namely the low-level processes that serve as the backbone for higher-level data analytics.
This infrastructure comprises hardware and software. Hardware includes processors such as CPUs, GPUs and TPUs (central, graphical and tensor processing units), transient memory and permanent storage, communication elements, powering and cooling elements, etc. Software infrastructure enables data access and pre-processing, programming, monitoring of data streams or hardware, etc.
Improving infrastructure for big data therefore requires advances in both hardware and software. While industry drives progress in hardware, academic research contributes significantly to new software that enables faster and more efficient processing of large datasets.
This project devised novel ways to analyse performance in cloud data centres, an important task in managing computing resources efficiently while minimising energy consumption.
This project extracted inferences from a network and explored graph analytics on different platforms – including combinations of in-core and out-of-core processing.
This project made several advances in the analysis of large graphs (networks) using In-network computing, namely the processing of data while in transit and before storage.
The aim of this project was to devise new techniques for the automatic or semi-automatic integration of data.
The aim of this project was to improve combinations of programming languages and databases. Results were integrated in Scala 3.
The aim of the project "Big Data Monitoring" was to develop algorithms that continually monitor incoming data for rule violations.
