Cyber Times International Journal of Technology & Management

Since technological limitations restrict the throughput of traditional single processing unit based sequential computers. High performance computers are increasingly using multiple processors. These processors work in parallel on different parts of the problem. This is referred to as parallel processing.

Parallel processing increases the responsiveness, speed and throughput of computers. It can solve bigger problems (scale-up) and solve existing problems faster (speed-up) than what is possible with conventional computer architectures. Leveraging on state-of-the-art VLSI technology, reconfigurable parallel processing architecture, and progress of parallel algorithms, C-DAC delivered  supercomputing solutions to end users covering hardware, systems software, and applications. PARAM is the first series of high performance computers from C-DAC. Using over 250 processors, these machines can provide seamless, scalable performance exceeding 1GFLOPS (peak) and over 3000 MIPS (peak). This paper aims at taking review of the architecture and applications of supercomputer PARAM, since the supercomputing is reshaping the traditional science and engineering disciplines. A new era of computational science has begun through the synergy of  supercomputers and sciences.