Funding: Polish Ministry of Science and Higher Education
Project period: 2021 - 2024
Project leader: Dariusz Plewczyński
Institutions:
Faculty of Mathematics and Information Science, Warsaw University of Technology
Dr hab. Dariusz Plewczyński, prof. universities from the Faculty of MiNI WUT obtained grant from the Ministry of Science and Education for the implementation of the knowledge system "Artificial intelligence system for the interpretation of the DNA sequence of the human genome". Funding in the amount of PLN 2 million was granted for the purchase of a cluster of three modern supercomputers NVIDIA A100 with a capacity of 5 PFLOPS each combined with 1PB modern data storage system (DDN). The computational resources are for the scientific activities of Laboratory of Bioinformatics and Computational Genomics http://genomics.mini.pw.edu.pl at the Faculty of Mathematics and Information Sciences of the Warsaw University of Technology. The goal is to train state-of-the-art deep learning algorithms for analysis and the interpretation of the DNA sequence of the human genome.
In May 2020, NVIDIA, the world leader in the production of computing cards widely used as an essential tool in machine learning, deep learning, genomics and many other research, introduced new computing processors codenamed A100 with more computing power and a better power-to-power ratio than previous solutions. NVIDIA DGX A100 https://www.nvidia.com/en-us/data-center/a100/ of the latest generation is a universal system for all applications related to artificial intelligence (AI) and the use of graphics processing units (GPUs) for computing and simulation biophysical (polymeric). The supercomputer integrates eight of the latest advanced NVIDIA A100 Tensor Core graphics processors, providing the first AI system with 5 PFLOPS performance and at the same time the first infrastructure in history that can be used both for data preparation and analytics, deep learning and inference, and biophysical simulations. The project will create a unique, universal, fast, scalable and easy-to-implement infrastructure for artificial intelligence to interpret the DNA sequence of the human genome on a population scale.