Chennai, October 2020: The Department of Computer Science and Engineering of IIT-Madras recently announced that it had developed a microprocessor christened ‘Moushik’. It uses a 180-nanometer semiconductor chip and can be programmed for a number of uses such as in credit and debit cards, surveillance cameras, consumer electronics, voting machines, and various sensors.
One instinctive reaction to this is: isn’t this re-inventing the wheel? Who needs a desi microprocessor when the stuffis available cheap and aplenty in the world?
Such a reasoning is totally wrong, says Prof V Kamakoti, who headed the team that developed the microprocessor.
In a chat with Business Line, Kamakoti pointed out that imported microprocessors are a potential security risk. “We know what it (the imported microprocessor) can do, but we cannot tell that it will not do what we don’t want it to do.” In a sense, an importer microprocessor is like a “black box”; we do not know what all it contains, whereas an indigenous one is not.
Kamakoti is also a member of the National Security Advisory Board. As India is about to move into a higher level of electronic gadgetry – an explosion of sensors is anticipated, with use in healthcare, precision agriculture and internet-of-things – and the role of microprocessors increases correspondingly. There are going to be billions of devices, and each would need a microprocessor.
Also, when uses imported processors are used, we will need to develop applications according to the processor’s capabilities. “This is like cutting the foot to fit the shoe,” says Kamakoti. Furthermore, in cases where a processor needs to be used for strategic application, even the security clearance will take a very long time – even as long as 15 years. And, if the seller of the processor stops supporting it, then you would need to buy a new processor and reconfigure the entire system for compatibility.
However, the Moushik microprocessor is completely open-source, which means the source codes are available in the public domain. Anybody could develop it further to suit his needs, without having to disclose it.
“We want start-ups to take up our microprocessor and create businesses out of it,” said Kamakoti.
What about Moushik having been built with ‘180 nm technology’?
The ‘nanometer’ in the context of semiconductors means the size of the transistors, which are like ‘brain cells’. The smaller the transistor, the smaller the device can be. Today, in a world where semiconductor conductor manufacturers such as Samsung and TSMC have gone as small as 5 nm transistors – leaving even Intel behind at 14 nm – 180 nm seems ancient.
To this question, Kamakoti observed that there are still thousands of electronic devices that work perfectly well on ‘180 nm technology’. India’s semiconductor laboratory in Chandigarh is fully capable of making the 180 nm semiconductor – so the Moushik microprocessor syncs well with the Atmanirbhar Bharat initiative. Size is not so critical when you talk of devices such as surveillance cameras or washing machines.
If India can make good with indigenous chips for a large part of its market, it can better exercise its purchasing power in the international markets, said Kamakoti.
What about costs?
Pointing out that the cost related to scale, Kamakoti said that if one million chips could be manufactured, they could be sold for 50 cents a piece, the same prices as imported. The Moushik is the third chip of the ‘Shakthi family’ that the CSE team in IIT-Madras has developed.