Centre for High Performance Computing

Our project, Dhakshina, is an answer to critics who have pointed out that FFSC has many dream projects, none of which have actually been implemented. Hats off to our excellent team who have worked day and night to make this dream come true as envisioned by our founders and supporters.

Our High Performance Computing Centre, which aims at building a supercomputing lab working at several gigaflops of speed is nearing completion. With focus on high-load applications involving 3D rendering and heavy calculations, this state-of-the-art Lab encourages students and staff interested in research-oriented activities.

This project is one of the first of its kind in the state of Kerala.

What is Dhakshina?

Dhakshina Cluster Series II is a High Performance Computing System developed by the FISAT Free Software Cell using the Beowulf Architecture.

Dhakshina started as a simple desire for improved performance of applications over that provided by a single computer. Typically much more cost-effective than single computers of comparable speed or availability, clusters have proved to be efficient in all regards for large-scale applications. Aimed at professionals and students interested in the research and development of parallel processing applications, Dhakshina is benchmarked at a top speed of 180 Gflops.

The Specs

The backbone of Dhakshina is formed by two IBM X series servers with Quad Core Xeon processors. It has 32 computational nodes, 3 login nodes, all Intel Pentium 4 machines (3.0 GHz with HT technology) with 1.23GB of RAM, and 80 GB of hard disk storage.

The cluser nodes in the Dhakshina are interconnected using Gigabit Ethernet LAN. The Gigabit ethernet enables the use of both the optical fibre cable and also the copper cables.

Dhakshina started as a simple desire for improved performance of applications over that provided by a single computer. Typically much more cost-effective than single computers of comparable speed or availability, clusters have proved to be efficient in all regards for large-scale applications. Aimed at professionals and students interested in the research and development of parallel processing applications, Dhakshina is benchmarked at a top speed of 180 Gflops.

http://chpc.fisat.ac.in