About – Exa4mind – Extreme Analytics for Mining Data spaces

A leap forward to

common European Big Data
and Extreme Data techniques

common European Big Data and Extreme Data techniques

in science, industry and SMEs

Why

EXA4MIND?

In an age of exponentially growing volumes of data, delivered in vast quantities by business platforms, sensors, simulations, satellites and the media, massive high-performance data analytics are in continuous development. Meanwhile, Europe’s most powerful supercomputers are mostly used for simulations, and efficient data access is limited to cluster file systems. Their potential power for use in many areas of modern data analytics remains untapped, whilst simultaneusly industry, scientists, and society often struggle to benefit from data pools.

As a co-design collaboration of universities, supercomputing centres, and application case providers from science, industry, and SMEs, the EXA4MIND project addresses scientific and technological challenges. We aim to effectively democratize access to and enable connectivity across EU supercomputing centers to achieve innovative solutions to complex everyday problems and to address critical challenges like data analytics, Machine Learning and Artificial Intelligence at scale.

We will build an extreme data platform that brings together large scale data storage systems and powerful computing infrastructures by implementing novel automated data management and effective data staging. This will enable integration with extreme data sources and support complex and advanced data analysis pipelines for knowledge extraction.

The

Challenges

Data Analytics, Machine Learning (ML) and AI at Scale, exploring the potential of an Extreme Data Database and analytics on Supercomputers, with AI & ML, data collection from heterogeneous sources, automatic data harmonisation and annotation on the fly, and data transport over high-end European networks.

New Solutions to Complex Everyday Pro- blems in Science, Industry & Society, for example, conceptual improvements of the methods used for optimising molecular-dynamics simulations, based on a massive simulation and observation database of 100TB growing to 1PB.

Going beyond the Limits of Today’s ML/AI, leveraging Supercomputing power to annotate driving scenes and explore cases of failure in Advanced Driving Assistance Systems (ADAS), to accelerate European ADAS development with automotive camera and sensor data that is growing from the PB to the multi-PB range.

Easy Access for SMEs to AI, Machine Le- arning (ML), and Supercomputing on Big Data including Satellite Imagery, enabling SMEs and industrial/societal stakeholders to leverage top-level supercomputing and networking infrastructures.

Advanced Queries and Indexing across Data Backends, leveraging the experience of universities and computing centres with Database Management System (DBMS), object, and file storage to offer high-performance unified indexes and standard queries to a range of back-end systems with a “technology enabler” approach.

Large-Scale Connectivity across European Supercomputers, Databases, and Europe- an Data Spaces, boosting interoperability between supercomputing facilities, data systems, and European Data Spaces as well as European Open Science Club (EOSC).

EXA4MIND offers a platform for Extreme Data that enables advanced data analytics on supercomputers and automated data management with support for integration with EOSC and European Data Spaces by design.