| Abstract [eng] |
The growth of electricity consumption raises problems related to electricity production and imports (when there not enough internal resources in the country). The majority of new technologies require larger and larger quantities of electricity. It can be seen everywhere, but especially the problem highlights when speaking about the modern information computer technologies. Newer devices are released and consume less electric power, but the amount of such equipment is increasing, which increases consumption of electric power. Consumption growth also concerns the distributed and parallel computing. In the world, every year, the number of supercomputers, clusters, grid and cloud computers is growing up. All these resources require a lot of electricity. Same applies to the most powerful computer in the world, \"K computer\". It requires 12 659.89 kW of electricity when running at full power. That much electricity would be sufficient to maintain about 60 000 usual households. Parallel computing has been developed and evolved rapidly during the last 30 years. Various technologies of parallel computing have been implemented. All technologies are different and require relevant knowledge from modern scientists. Scientists can use different technologies: supercomputers, cluster systems, grid’s, cloud computers and other computing resources. Depending on the technical implementation of computing resources, they may be provided with different software. New libraries are created as well as new programming standards for parallel computing. For example, MPI, OpenMP, gLite, XMPP-MPI, CUDA, OpenCL and other parallel programming standards have been implemented. When talking about parallel computers, most often we have in mind several personal computers connected by a network into a single computing resource. If the resource is small, then its computational power is limited. If the computational power is growing up, so does the number of computers of a computing resource. This has influence also on electric power consumption. At present, it is a big problem that requires attention from scientists and manufacturers of computer equipment. In 2005 this problem was named - Green Computing. San Murugesan was the first to define the term of green computing as “the study and practice of designing, manufacturing, using, and disposing of computers, servers, and associated subsystems – such as monitors, printers, storage devices, and networking and communications systems – efficiently and effectively with minimal or no impact on the environment.” At the present time, we can also find other names of this problem: green information technologies (green IT), environment-friendly IT communications and computing. The topicality of this problem is stressed by the list of most efficient energy aware supercomputers named - Green500 (existing since 2005). Optimization methods can help to solve the electric energy consumption problems. In addition, solution of this problem can minimize the use of available computing resources to reduce their electricity consumption. This fact is emphasized by the green computing expression – “Higher computing power with the smaller influence on the environment.”. |
