• Heterogeneous computing cluster HybriLIT. HybriLIT is a unique thing in the multifunctional center for data storage, processing and analysis of LIT JINR. Heterogeneous structure of the cluster HybriLIT allows carrying out computations with the use of various parallel programming technologies such as MPI, CUDA, OpenMP, OpenCL.
  • HybriLIT cluster includes 10 computational nodes, nodes with GPU NVIDIA Tesla K40 and Tesla K40, a node with Intel Xeon Phi 7120P coprocessors, and a node containing NVIDIA Tesla K20x and Intel Xeon Phi 7120P coprocessor. Each computational node includes two Intel Xeon processors.
  • HybriLIT provides users with the following services: ¤ Indico - a system for organization and execution of events, conferences and lections: http://indico-hybrilit.jinr.ru/ ¤ FREE IPA - a service for user account management: http://ipa1-hlit.jinr.ru/ ¤ Git - a system of file version management: https://gitlab-hlit.jinr.ru/
  • HybriLIT group developed information-computer environment that includes specialized services and that offers a wide range of opportunities on the analysis of new approaches and technologies of parallel programming, convenient tools for application development, adjustment, profiling and for carrying out resource-intensive computations connected with mathematical support of researches held at JINR.
  • HybriLIT cluster includes 10 computational nodes, nodes with GPU NVIDIA Tesla K40 and Tesla K40, a node with Intel Xeon Phi 7120P coprocessors, and a node containing NVIDIA Tesla K20x and Intel Xeon Phi 7120P coprocessor. Each computational node includes two Intel Xeon processors.
  • Regular tutorials and trainings on the parallel programming technologies oriented on users with different level of proficiency in the field of heterogeneous computations are being held on the basis of the heterogeneous cluster.
  • Heterogeneous cluster HybriLIT
  • Heterogeneous cluster HybriLIT
  • Heterogeneous cluster HybriLIT
  • How to work on a cluster posted in the Users

 

“HybriLIT” heterogeneous cluster is a computation component of a multifunctional center for data storage, processing and analysis of the Laboratory of information technologies JINR; and it is intended for performing computations with the use of parallel programming technologies. Heterogeneous structure of computational nodes allows developing parallel applications for the solution of a wide range of mathematical resource-intensive tasks using the whole capacity of multicore component and computation accelerators: Nvidia graphic processors and Intel Xeon Phi coprocessors.

Registration Indico Git Lab User Guide