GUJCOST has setup a Supercomputer facility – PARAM Shavak, at Computer Engineering Department of Government Engineering College-Rajkot in December 2017. To contribute to the research and development of India, C-DAC has developed a low cost and an energy efficient solution, which is as good as a supercomputer in a box namely – PARAM Shavak DL (Deep Learning) GPU. PARAM Shavak DL GPU, has been equipped with advanced technologies to perform the high-end computations for the scientific, engineering and academic community.
PARAM Shavak will be able to address and catalyze the research using artificial intelligence (AI), Deep Learning and promises to be a boon to the academic and scientific community. Additionally, it will help in creating a workforce that is aware of AI skills (capacity building) and promote research and teaching by integrating leading-edge emerging technologies.
Processor: Intel Xeon E5-2680 (14 cores each of 2.4GHz)
RAM: 64 GB ECC (8GB x 8) DDR4
HDD: 8 TB
OS: Ubuntu 18.04 LTS
Frameworks : Torch 7, Caffe, TensorFlow, Keras, MXNet, Theano, Paddle
Performance Libraries: CUDA Toolkit, cuDNN, cuSparse, cuBLAS, TensorRT
Visualization Tools : DIGITS, TensorBoard