The NHPCC#
Introducing the plan#
The National Cloud Computing Network was opened in 2009 with the aim of providing complete and comprehensive computing services to the industries and scientific and research centers of the country and under the support of the Presidential Vice President for Science and Technology.
Isfahan University of Technology, as one of the several supercomputer centers in the country, at that time, implemented a part of this national plan in the form of the establishment of the Sheikh Bahai National Supercomputing Center.
Currently, The Sheikh Bahai National High Performance Computing Center (NHPCC) is a unique center in the country, which is able to meet a wide range of computing needs of advanced industries and scientific and research centers of the country by using the power of expert and experienced forces as well as providing advanced hardware and software infrastructures.
The planned arrangements in the Sheikh Bahai NHPCC are such that comprehensive and complete hardware and software facilities as well as technical and specialized educational and consulting services are available for optimal use (on site or remotely). Applicants can use this center.
Hardware features#
The Rakhsh Supercomputer#
- Total processing power of more than 2 Petaflops
- 6592 XEON and AMD processing cores
- 40 Gbps Infiniband connectivity network
- 96256 graphics processing cores
- Has more than 30 terabytes of main memory
- 300 terabytes of information storage capacity
- Designing, compiling and developing the first comprehensive management and user portal of supercomputer centers
- Designing, compiling and developing the first scientific and specialized virtual forums
- Support for a wide range of computing software
- Providing cloud computing service
- Connected to the national network of Turin
The Panthera Cluster#
The Panthera
Panthera is named after one of the scarce animal spieces in Iran which was extinced a hundred year ago. But now he has come back again and roars like a true lion.
It has 84 processing nodes with a processing power of 2.24 Petaflops 5632 AMD processing cores Has more than 20 terabytes of main memory Gigabit communication network It has a local storage space with a capacity of more than 60 Terabytes Connect to central storage with 200 TB of space Connected to the Turin national network.
NHPCC compute nodes#
Currently there are for types of nodes we have clasified based on their CPU Architectures:
| Common Name | Architecture | CPU Cores | CPU Type | GPU | Memory |
|---|---|---|---|---|---|
| EN NODES | AMD EPYC 7542 32-Core Processor | 64 | real | - | 1 TB |
| EN NODES | AMD EPYC 7502 32-Core Processor | 128 | thread | Tesla V100 | 1 TB |
| XN NODES | Intel XEON 5672 | 48 | thread | - | 96 GB |
| CN NODES | AMD Opteron™ Processor 6174 | 48 | real | - | 96 GB |
below are the total number of nodes at NHPCC:
| Common Name | Num |
|---|---|
| EN NODES | 19 |
| XN NODES | 32 |
| CN NODES | 64 |
Our GPU Cards
It is worth mentioning that currently we have 16 Tesla V100 Cards Actively working inside Panthera's GPU nodes.
Software facilities and computing services#
The NHPCC offers a wide range of services to users and subscribers. By using several clusters with different computing powers and using the latest processors in the world, this center can respond to many needs of users at different levels. In order to facilitate the use of the center's computing services, a wide range of computing software of various fields of engineering and science have been installed and launched on the Supercomputer, among which the following applications can be mentioned:
- Computational fluid mechanics and dynamics
- Meteorology and nuclear science
- Aerospace, oil and gas, automobile and defense industries
- Mathematics, chemistry and computational physics
- Computational nanotechnology and bioinformatics
What services we provide#
Educational Services#
The NHPCC Supercomputing Center organizes a wide range of educational services on a periodic basis and with prior notice or on request (at the request of subscribers). For example, some of the educational services and specialized workshops of this center include:
- Practical training and Linux management (introductory, intermediate and advanced)
- Specialized workshops for advanced programming in Linux
- Specialized workshops for working with SLURM
- Specialized parallel programming workshops on supercomputer (CUDA, MPI, OpenMPI, OpenCL)
- Workshops for working with specialized software
Consulting Services#
The use of parallel processing hardware and software and cluster systems requires familiarity and mastery of various scientific and technical topics. The experience of the collaborative team in the management and administration of the Sheikh Bahai National HPC Center and our many years of experience in developing and using parallel processing hardware and software systems has made it possible for us to be able to provide a wide range of users and applicants to use supercomputers and offering consulting services.
Cluster Hosting#
The installation and operation of a cluster system should be done in a suitable environment in terms of safety system, power supply, cooling, emergency power, etc. This requires relatively heavy initial investment and costs, as well as ongoing maintenance costs. The National Supercomputing Center is able to host the cluster systems of other institutions and organizations by taking advantage of the suitable and large space and the necessary infrastructure based on up-to-date standards. This has not only led to significant savings for organizations, but it will also be very effective in the professional support and maintenance of their equipment.
Cluster design, installation and commissioning#
The National Supercomputing Center provides complete services in the design, installation and commissioning of cluster systems. The collaborative team in the management and administration of the center, taking advantage of the most practical and successful experience in the country in the field of designing, installing and setting up cluster systems, can design an optimal cluster system with minimum cost and in the shortest time by assessing and identifying your computing needs. and by presenting the results of the system performance tests, install and operate the system in the desired location of the applicants. Also, this center can undertake maintenance and post-delivery services. Users who order their cluster systems through this center can benefit from the consulting and training benefits of the center with special discounts.