AI & Distributed Cloud, the other path
cleared by Hivenet & Inria

Hivenet and Inria have strengthened their partnership around artificial intelligence and the distributed cloud, as a strategic alternative to centralised solutions. Initially focused on distributed data storage and large-scale file sharing, this public-private collaboration aims to explore new frontiers by moving away from traditional data centre infrastructures.

Over the past two years, Hivenet and Inria have made significant advances in optimising redundancy and data placement algorithms, improving the availability of files on a decentralised network. These innovations are now at the heart of Hivenet's infrastructure, which has been adopted by more than 325,000 users in 190 countries.

This strengthened partnership has been structured around 3 strategic axes:

1/ developing AI models capable of operating efficiently on different types of hardware

2/ developing advanced encryption techniques meeting the needs of the most demanding sectors

3/ continuing in exploring new methods for large-scale distributed computing, in particular to take advantage of computing resources made available intermittently.

A strategic public-private partnership

Since 2023, Inria and Hivenet have been working together as part of the ALVEARIUM challenge, to develop distributed storage and computing technologies in order to reduce hyperscalers' dependence on cloud infrastructures. This approach has been based on the use of decentralised computing resources, mobilising the computing power available on different types of hardware, from personal PCs to high-performance servers. With this in mind, a second challenge, the CUPSELI challenge, was officially launched earlier this year at last WAICF.

This major scientific collaboration mobilises a large-scale scientific community, bringing together 11 Inria research teams - ARGO, COAST, COATI, OCKHAM, MAGELLAN, MIMOVE, NEO, TADAAM, TOPAL, STACK and WIDE - from six research centres: Rennes, Bordeaux, Lorraine, Côte d'Azur, Lyon and Paris. The project is part of a close collaboration with leading academic institutions, including the University of Bordeaux, the University of Lorraine, the University of Rennes and the Côte d'Azur University, as well as research bodies such as the CNRS, the Bordeaux Polytechnic Institute, the Institut Mines-Télécom and the École Normale Supérieure de Lyon.

This exploratory work will be tested on cutting-edge infrastructures, in particular the Jean Zay and SLICES computing platforms, guaranteeing concrete and operational applications of the research results.