• July 11, 2021

How to make a ‘supercomputer’ using electric appliances

India is a land of plenty, but for a few hundred people in the country’s north, life is very different.

Indias most populated area is the Andhra Pradesh state of Telangana, home to the capital city of Hyderabad.

There, the countrys first supercomputer, called ‘Kolpana’ or ‘the machine’, was built.

The system was designed to run at 1.3 petaflops, or about 3 million trillion calculations per second, which is more than three times faster than today’s most powerful supercomputer. 

Indias largest internet service provider, Reliance Communications, has partnered with Microsoft, IBM, and Intel to make the system, which was built by a team of engineers, developers, and students at the Indian Institute of Science (IISc), Hyderabad, India.

“The team at IISc was not only responsible for designing the supercomputer but also built the software that powers it,” said Vikas Prakash, director of IIScs computer science department. 

“The design and development of the supercomputing hardware were done by an engineering team at Microsoft and Intel. 

We are very grateful to Microsoft for their support of the project and for allowing us to use their technology,” said Rajesh Kumar, president of the India Supercomputer Center, a non-profit organisation. 

According to Microsoft, the super computer was built for use by the Indian government, the National Highways Authority of India, and private enterprises. 

The supercomputer uses eight 8-core Intel Xeon processors running at 2.5GHz and 16GB of DDR3 RAM, running on a 64-bit, 32-bit and 64-core architecture. 

In addition to its primary purpose, the IIS supercomputer is also designed to make its users aware of the benefits of artificial intelligence. 

For example, the software used to run the system was built with deep learning.

This technology can automatically identify patterns and patterns in the data it processes. 

IISC’s supercomputer could potentially make it possible for a person to go to their local supermarket and buy an extra packet of groceries, or get an extra package of coffee, as a result of using machine learning. 

Image credit: Microsoft