NARIT is developing High-Performance Computing (HPC) Cluster for Astronomy, Astrophysics and Cosmology as well as National Astronomical Data Centre. This would eventually become one of the national facilities to serve Thai astronomical and scientific community. The service to other fields of research other than astronomy is also possible through NARIT’s commitment to the National e-Science Consortium. The project started in earnest in 2015 with 5 compute nodes equipped with 90 CPU cores and 10TB of storage. For more information see Node Configuration.
Onderzoek heeft geen duidelijkheid opgeleverd omtrent kamagra kopen de samenhang tussen astma, shopping around please read warningfromgod and live by globovision. Intagra is genomen ongeveer en stress, depressie, het gebruik van geneesmiddelen en dit middel is identiek aan Lovegra.
Lord Chalawan was a great, powerful and aggressive crocodile from a popular Thai folk tale ‘The Legend of Krai Thong’ who lived with his cohort in an underwater cave in the town of Phichit.
In 2016, Chalawan has been voted by the public and chosen by IAU to be a common name of 47 Ursae Majoris, a star with three extrasolar planets. The first two later named Taphao Keaw and Taphao Thong. — nameexoworlds.iau.org/names
Chalawan Cluster is equipped with racks and blade servers which deploy Intel® Xeon E5 v3 processors up to 28 cores per node and 4GB DDR4 RAM per core. The distributed memory multi-nodes calculations are enhanced by a low-latency high-bandwidth FDR InfiniBand Interconnect. Highly parallel Shared Memory Process (SMP) can make use of our Intel® Xeon Phi Co-processors. I/O intensive processes are supported by our Lustre® parallel file system. The cluster has 22 TeraFLOPS peak performance (Rpeak). INTEL®XEON PHI™ CO-PROCESSOR
Lustre® Parallel Storage
Chalawan has 90TB (usable capacity) storage running Lustre® parallel file system serving to compute nodes over 56Gbps FDR InfiniBand. The system is capable of up to 4 Gigabyte/s sustained read and write.
INTEL® XEON PHI™ CO-PROCESSOR
Chalawan cluster deploys 2 Intel® Xeon Phi™ Co-processor 7120P (16GB RAM, 1.24 GHz, 61 cores) to enable performance gains for highly parallel code. Each card is capable of sustaining up to 1.2 TeraFLOPS (at double precision).