The Super Computer Using 1544 GPUs of NVIDIA Tesla X2070
GPU computing module has proved itself more than just a competent asset that is used in supercomputing systems. NVIDIA Tesla is now seems to be a little proud, the article of the recent components in its ranks has managed to get a place to be used as part terhandal supercomputer owned by a Russian tycoon.
GPU computing module has proved itself more than just a competent asset that is used in supercomputing systems. NVIDIA Tesla is now seems to be a little proud, the article of the recent components in its ranks has managed to get a place to be used as part terhandal supercomputer owned by a Russian tycoon.
Some of the problems of humanity are often faced with the problem such as disease, environmental degradation, and the like, can basically be solved by the ability of an increasingly powerful supercomputers. Although so far the success rate obtained varies greatly, but there was no mistaking the fact that high-performance computing clusters will be able to contribute to ongoing and contribute to human progress.
Recently, GPU computing modules, such as AMD and NVIDIA Tesla FireStrem series, has proved that the components diusungnya able to perform specific tasks in supercomputing, parallel processing in particular, at a very high level of performance. And based on recent news circulating, even the ranks of the GPU will now be used to upgrade the HPC systems at the Lomonosov Moscow State University, Russia.
"Our research requires computing resources that are very large, and we need to give this performance as efficiently as possible," said Victor Sadovnichy, one of the academics and the Rector of Moscow State University.
The only way for us to achieve the various objectives of the existing system with GPU / CPU-based hybrid".
This system has been used to study global climate change, ocean modeling, the formation of galaxies and the discovery of drugs for medical purposes.
Once the cluster is fully upgraded, the system will have a 1544 NVidia Tesla GPU X2070, and one GPU for each quad-core CPU in the installation. Peak performance of 1.3 petaflops itself into, and this is number one in Russia and even among the fastest in the world.
Some institutions such as the Keldysh Institute of Applied Mathematics, Lobachevsky State University of Nizhni Novgorod (NNSU) and the Scientific and Educational Center of Parallel Computing at Perm State University reportedly is also going to use the NVIDIA GPU module in its projects.
"There is a potential shock to the system solution based hybrid GPU / CPU is to help us cope with a large number of scientific challenges such as learning a survival system, bio-photonics and computational mathematics," said Victor Gergel, Dean of the department of computational mathematics and cybernetics at NNSU, director of the Scientific and Research Institute of Applied Mathematics and Cybernetics.
"In cooperation with NVIDIA, NNSU able to give more students and researchers Us to access to computing resources that will significantly increase the speed of their work".