THE PROGRAM “SKIF”: TO BE CONTINUED

At the past meeting that took place in Minsk the Council of Ministers of the Russian-Belarusian Union State approved the scientific-technical program “Development and utilization of the GRID-technology soft hardware and advanced high-performance (supercomputer) systems of “SKIF” range” for the period of 2007–2010
The head executor on the part of Belarus is the United Institute of Informatics Problems of the National Academy of Sciences of Belarus (UIIP NASB). The Head of the State Supercomputer Shared Centre of the United Institute of Informatics Problems of the NAS of Belarus Vitaly Stetsurenko told about the actuality of the forthcoming researches.
— Vitaly Ivanovich, what objectives do the executors of the new program face?
— In the course of the program’s implementation we will carry on the studies that were initiated within the framework of “SKIF”, a program of the Russian-Belarusian Union State. The program was completed in 2004 and, as you know, its main result was generation of the Russian-Belarusian supercomputer. The evidence of the success of the program is the fact that in November 2004 the top-of-the-line model of “SKIF” range — cluster “SKIF K-1000” — took the 98th line in the list of TOP-500 of the world’s most powerful computer installations. However, the progress, especially in the IT field is not at stop, and now we have to maintain the headship gained.
— For now the term “GRID-technology” is familiar only to specialists. What does the innovation look like?
— GRID-technology is the technology of generating the effective geographically-distributed networks, which help to join computers to various hardware and software systems. GRID-technology can give occasion to generation of super powerful information-computer systems which combine low operating costs with high performance at the expense of raising the utilization rate of all GRID-network resources. The innovation makes it possible to process large volumes of dataflow at a higher level, to analyse them on the basis of series of criteria, to simulate necessary situations on the remote supercomputers as well as to use complex business applications of huge amount of calculation. Thus, GRID-technology can help to solve scientific, scientific practical, engineering, medical and social problems of high degree of complexity that were out of question before. Nowadays state GRID-projects are being developed practically in every advanced country. For instance, in Europe all GRID-technology activities are coordinated under Data GRID project (9.8 billion have already been invested in the project); in the United Kingdom a special state body has been established to develop GRID-systems. According to the forecasts of the western experts the volume of investments in GRID-projects over the period of 2003–2008 will increase twenty times. As GRID-projects are still in the early stage of their life history — namely, at the scientific stage — Belarus and Russia have chances not to lag behind in this field.
— And what are the tendencies of the development?
— Firstly, we have to further develop and manufacture high-performance computing. In particular, we are planning to create a number of different-purpose systems based on the new generation supercomputers of “SKIF” range. Secondly, the specialists have the task to diminish dependence of Belarus and Russia on utility supply from abroad and on ready-for-service high-performance computer systems as well as to reduce risks in the field of information security — as they arise while using imported software and hardware for high-performance calculations.

Vladimir Dorokhov
Заметили ошибку? Пожалуйста, выделите её и нажмите Ctrl+Enter