The Cyberscience Center, Tohoku University, the Cybermedia Center, Osaka University, National Institute of Informatics and NEC announced the successful demonstration of one of the world’s fastest vector supercomputing environments by creating a single virtual system through the connection of two remotely located vector supercomputers on NAREGI (National Research Grid Initiative) middleware developed by NII.
Vector computers are suitable for carrying out large scale scientific computing such as fluid dynamics, structural dynamics computations, new material research and climate simulations with high computation efficiency, as well as being an important base for cutting-edge R&D and product design. Tohoku University has deployed 16 nodes of NEC’s SX-9 supercomputer (maximum vector theoretical performance: 26.2 TFLOPS) and Osaka University has deployed 10 nodes of the SX-9 (maximum vector theoretical performance: 16.4 TFLOPS), each of which boasts a high speed connection to the other through SINET3 (Science Information NETwork 3).
NAREGI middleware enables large-scale computing resources at research and development centers scattered over a large area to be closely interconnected through high speed networks. These network connections can be viewed as a single massive virtual computer that efficiently implements large-scale parallel simulations, which were formerly difficult for individually isolated computer systems to carry out.
A new Grid middleware component, the “GridVM for the SX Vector Computer,” was developed by enhancing the existing capabilities of the NAREGI middleware, such as job management, information provision and resource usage control. The enhanced GridVM maintains high compatibility with the local job scheduler (NQS) on the SX-9, which enables the efficient use of vector computing resources even in the Grid environment. Moreover, it permits the co-existence of conventional (non-Grid) jobs and Grid jobs, allowing the computing center to provide a pioneering new cloud-computing service.
In this demonstration experiment, a parallelized electro-magnetic field simulation program was run by interconnecting two SX-9 systems both at Tohoku University and Osaka University with the use of parallel programming libraries for shared memory and distributed memory. As the first step in establishing the cloud computing environment, the virtualization of the computing resources at both centers was made by integrating the newly developed GridVM for the SX-9 into the NAREGI middleware. Furthermore, it was successfully demonstrated that running jobs is possible with the automatic, selective allocation of jobs between two supercomputers in accordance with the system load status of each system, making the maximum utilization of resources over the entire Grid possible.
These organizations will continue their efforts to realize a vector-based cloud computing environment as a new academic information infrastructure that allows the overall application software to run efficiently with enhanced usability and reduced cost through cooperation with many organizations that possess vector computers.
As a result, it is expected that an advanced scientific computing environment can be established with the following new services: running jobs through the automatic assessment of user requirements, automatically searching for application locations in accordance with user preferences, automatically detecting the most effective computing resources, providing the user with high performance computing services and university-created application software services made available in a unified environment via the consolidation and virtualization of computing resources over the Internet.