Last Updated on 20 November 2022
DATA-INTENSIVE COMPUTING PLATFORM
Head of the platform : 
High-performance computing unit of the University of Medea offers to the scientific research community, affiliated with the University, powerful tools and software necessary to carry out complex scientific or technical applications in various academic and industrial fields.
Data-intensive computing or High Performance Computing is a field where equipment and software necessary for the execution of complex programs and technics that are used in various academic and industrial fields, like the hydrocarbon industry, pharmaceutical industry, medical imaging, and meteorology and physics simulations.The progress of simulation and computing technics used in these fields, along with the development of computing capacities and parallel computer architecture, allows achieving a computing speed that have never been expected and solving huge digital problems in an increasingly less time.Some parallel computer infrastructures have achieved, nowadays, a processing speed of Petaflops, i.e.,1015 floating point operations per second. The processing speed is related to the quantity of calculation, which is the number of elementary operations, used per a unit of time. Therefore, to consider a software as intensive depends on its operation requires using a big number of processing resources so that it able to operate in a reasonable time in comparison with the human time scale. For instance, in meteorology, it would be too late to predict next week’s weather in a year’s time of processing.Fie
lds of application:Data-intensive computing influences various important fields that have direct impact on the economy such as petroleum industry, biotechnologies, pharmaceutical industry, medical imaging and meteorology…etc. All these fields need great computing capacities.Data-intensive computing platform of the University of Medea:The data-intensive computing platform project of the University of Medea aims at creating a center of expertise dedicated to scientific computing, drawing on the human and material resources available. It possesses a supercomputer having the latest processing technologies with a capacity of around 08 Teraflops, i.e. 8 billion operations per second. It will be a valuable asset for the research community at the University of more than 1000 Persons.
If fact, researchers from the University of Medea and from the economic partners used to suffer of an enormous slowness when using professional software programs for data processing. Nowadays, however, an operation that used to take 15 days is achievable in 30 minutes today thanks to supercomputers. Therefore, this solution offers a 100% of additional performance in comparison with the users’ needs.
Having this solution, researchers will not have to travel to carry out data computing which allows research organization to optimize their research costs. The University of Medea was the first Algerian university to have an IBM data-intensive computing unit. This is just the first phase of this project intended to be used locally. Future phases consist of the creating an HPC Cloud or Grid between HPC units of different Algerian universities.
Cluster composition:
Our cluster is composed of:
- 32 nodes.
- Each node is an IBM X3550 M3 sever with two processors.
- The processors are Intel Xeon X5675, 6 Cores, 3.06 Ghz, 12 M Cache.
- One administration server IBM X3550 M3
- One processing server IBM X3550 M3
- One storage server IBM X3550 M3 connected to a storage array IBM DS3512, 22 To.
Installed software and libraries:
-Operating system:
- Redhat 6.0 Enterprise Linux for HPC.
-Programming tools:
C, C++, Fortran…etc.
How to get access to the cluster?
- Submit the form approved by High-Performance Computing Unit located at in the Commun Services Building.
- You will receive the confirmation by e-mail of the creation of your account as well as the username and the password.
- In case you face any problems, please send an e-mail to
or go to the HPC Unit to discuss your issue with the HPC Unit responsible.
