Aug 25, 2024 | Posted by Abdul-Rahman Oladimeji
Six Russian supercomputing centers are teaming up to form a research consortium called the Distributed Scientific Supercomputer Infrastructure consortium. These facilities will combine a relatively meager 1.5 petaflops of compute performance with scientific data storage systems exceeding 15 PB across 900 servers. The six centers included in the consortium already support around 240 organizations across Russia, and contributed to more than 300 research papers in 2023.
“For various reasons, the time has come for systemic solutions in matters of building a national supercomputer infrastructure,” commented Alexey Anatolyevich Sorokin, PhD. “We need to move away from competition between shared use centers to combining their resources and competencies. Today, there is a demand for a distributed infrastructure that should cover technological sites in different regions of the country connected by high-speed communication channels, and equipped with modern computing systems with different architectures and specialized data storage systems.”
Nikolai Yuryevich Lukoyanov, director of the N.N. Krasovsky Institute of Mathematics and Mechanics added: “This project not only strengthens the interaction between various supercomputer centers, but also ensures more efficient use of high-performance computing systems in solving complex scientific and applied problems. The consolidated infrastructure will allow us to significantly expand the range of research, improve the quality of data processing, and support the development of science and technology in various regions of the country.
