Optimization of Applications on Supercomputers, June 20, 2024, Sozopol, Bulgaria

Optimization of Applications on Supercomputers, June 20, 2024, Sozopol, Bulgaria

  • Post category:Blog

On June 20, 2024, within the framework of the 4th international conference “Numerical Methods for Scientific Computations and Advanced Applications” (NMSCAA’24), held on June 17-21, 2024, in Sozopol, Bulgaria, a half-day tutorial session titled “Optimization of Applications on Supercomputers” took place. The session was organized by the team from IICT-BAS involved in the EUROCC2 project. Due to the considerable interest, online participation was also provided, with a total of 57 participants, including 40 onsite and 17 virtual attendees.

The tutorial was divided into two parts: (1) “Optimization of Applications on the HEMUS Supercomputer” and (2) “Cray: Programming Environment.” In the first part, Prof. Emanouil Atanassov (IICT-BAS) presented the characteristics of the HEMUS supercomputer, which was installed in October 2023 and has been operational since 2024. HEMUS is ranked 395th in the Top500 list of the world’s most powerful supercomputers as of June 2024. The presentation covered benchmark tests, analysis of results, and recommendations for optimizing applications. Special attention was given to the potential for collaboration with industry through a portfolio of services supported by diverse competencies in high-performance computing, high-performance data analytics, and artificial intelligence. The presentation was followed by practical demonstrations of how to work with the HEMUS supercomputer provided by assistant professor Svetlozar Yordanov (IICT-BAS), along with extensive questions from the audience.

In the second part of the tutorial session, the Cray programming environment was introduced by V. Saviak (Hewlett Packard Enterprise). The HPE Cray Programming Environment (CPE) offers programmers a comprehensive set of tools for application development, porting, debugging, and tuning. The programming environment simplifies the transition to new hardware architectures and configurations by automatically applying optimizations to HPC applications using existing programming models with straightforward recompilation.

The tutorial was concluded by discussions on the technical specifications of the HEMUS supercomputer, potential installations of various pieces of software that could benefit users from the academic sector and small to medium-sized enterprises.