India is not too late in quantum computing: Tata Institute of Fundamental Research's Rajamani Vijayaraghavan

As advances took place, tech behemoths like Google, International Business Machines (IBM) and Intel threw their hat in the ring with billions of dollars in investment. According to an estimate, China has invested $15 billion even as global funding in the technology has topped $36 billion.

At the New Delhi conclave last week, a senior defence official said the wars of the future will be fought on the quantum front. While sizeable quantum computers do not yet exist, the fear among governments and industry is that it will be a Y2K moment once the technology comes to life.

Apart from strategic affairs, quantum tech is also expected to play a central role in areas like drug discovery, financial modelling, weather forecasting and fighting climate change.

Moneycontrol spoke with Professor Vijayaraghavan, a quantum computing scientist at Tata Institute of Fundamental Research (TIFR) who returned to the country a decade ago after studying at the universities of Cambridge, Yale and Berkeley, to take stock of India’s progress in the area.

Edited excerpts: 

At the quantum tech conclave in Delhi last week, it seemed as though most Indian quantum start-ups are working on secure communications rather than quantum computing. Why so?  

See, in terms of the technology required for doing quantum key distribution (for secure communications), or associated areas like quantum random number generators, it is now fairly well developed internationally. Also, the essential tools and core pieces of hardware and things you need are relatively low-cost and easily available, compared to quantum computing.

If you shift to quantum computing, even if you just take a very basic estimate, it needs about a million dollars just to set up a very basic lab. This was the kind of funding that was given to me when I started in TIFR about 10 years ago, just to have a very basic setup and do simple experiments.

I'm not even talking about the resources you need to develop prototypes and technology. And even this $1 million is an underestimate, because there are other facilities, which were already existing in the institute for device fabrication, clean rooms, and other tools. If I count all of those, then you are talking about a couple of million dollars, at the very least, to get started.

Now, of course, a company could hire these resources from other places. But such facilities don't exist in India, and as of now, even internationally, they are limited, where a company with no infrastructure of their own, but just the idea and expertise could go to a place and say, hey, you know, here is my design, manufacture this for me, and we will put it together.

Experts say that India first missed the bus in taking a lead in semiconductor technology and now even in artificial intelligence. They say it is now or never if we want to be leaders in quantum computing. What is your take?  

I broadly agree with it, but let me expand on that. Before something becomes close to technology, there is decades of basic research where they are not even thinking about technology. I started my PhD in the year 2001, i.e. over 22 years back. And that was a turning point when the field of quantum computing was taking off. That was on the back of basic research in quantum computing from the 1980s.

It is absolutely true that if you want to do this, you have to do it now. And with this particular technology, we are not so far behind. Barring a few areas where things have sort of more or less been figured out, we still need several breakthroughs. Even though it looks like IBM has reached so many qubits (a qubit is the basic unit of information in quantum computing) and Google that many, they still don't have a useful quantum computer today. None of these quantum computers that exist anywhere in the world are really useful in the sense that they can replace something to do the work much better.

At how many qubits do you think we can solve real world problems better than existing supercomputers?

It really depends on what you mean by qubits. So there is this notion of physical qubits and logical qubits. The difference is that physical qubits are unstable. Unlike conventional bits in our regular computers, which are extremely stable, and you don't have to worry about them at all, these things are not stable. They actually randomly change their state and give rise to errors. And this happens over a certain timescale, which can be as short as 10s of microseconds or a few milliseconds.

Today’s quantum computers are essentially limited in how long you can use them, or essentially, what is the longest calculation you can run before you can't trust the results of what the computer is outputting. This is a fundamental problem across all quantum computing platforms today. There are some clear paths to solve this, but it is not clear which particular method would win.

And this is where I want to go back to my earlier point that India is not too late. If we invest now, we could create those breakthroughs that can make this technology useful. If enough people start working, and it's not just one or two people in the entire country, then you can hope that these breakthroughs can happen here. And then that will allow us to take a lead in some areas.

For superconducting qubit research, as of now, there are four academic labs in India. I was the first one who started about 10 years ago, and then two others started maybe about six and four years ago, and one has just joined.

In the US, there are at least 20 or more academic labs, in addition to Google, IBM, or national labs like Fermilab, and I'm not even counting theoretical groups. Many of them have been working for over 20 years or so. And you know, similarly, in Europe, there will be at least 10 to 15 labs just in superconducting quantum computing.

If you look into some of the other subfields like trapped ion quantum computing, it's even smaller.

Decisions about moving back to India can be very complicated, you know. There are all kinds of personal and professional reasons, but the point is very simple. You cannot attract someone to come to India to work in a field that requires a certain level of investment by telling them that we will not give it to you.

For me, I always wanted to come back to India and do something if it is possible. And once I finished my PhD, I started looking around and realised that there are lots of labs, which are fairly high-tech and at the cutting edge. Funding seemed to have improved in the previous few years or decade. It was important that I would be able to do the kind of research I was interested in doing rather than compromise. If that had not happened, I would have taken up a job abroad.

Has the funding scenario gotten better since then?

It's now becoming better, at least among the bigger institutes, but it's still quite a challenge. There is no real guarantee that you will get the kind of funding needed for experimental quantum research, and we are trying to solve this problem with the new funding programmes.

There is a reluctance to invest so much money in a new person. The idea of a startup grant is very clear in the West – your offer letter clearly says that you will get $1 million and you have to spend it in the next four to five years… everything is spelt out. We are not able to do that.  You have to keep your fingers crossed and hope for the best.

This actually does pose a barrier to entry because it's not competitive. If I have offers from an Indian institute, where everything is ambiguous, and some other foreign Institute, where everything is clearly spelt out, then it is very difficult to make a decision in favour of the former. Unless, of course, there are personal reasons to be in the country.

It should be a much more streamlined process for the government to disburse funds now as many new advanced research and academic institutes have come up. What are we missing here? 

Oh, no. It is not easy at all. The bigger institutes have some advantages in terms of flexibility of using their funds, but it is not easy to get big grants. Currently, the amount that you can get for a standard Department of Science and Technology (DST) project is Rs 10 lakh-50 lakh. That’s the typical range. Special programmes can provide bigger funding but they are rare.

Usually, it is easier to get big funding when a new institute is being built to set up new facilities. From thereon, things get tougher as you reach a steady state. Look, you need investment if you want to work in quantum tech. Of course, one can make a call that ‘Okay, we cannot afford this research, so we will not do it.’ But if you want to, there is no other option, you have to find ways to support this.

A few years back, the DST started a programme called QUEST (Quantum Enabled Science and Technology) where 50 scientists got funded with a combined funding of about Rs 200 crores. It was the first major effort in funding quantum tech research where an individual group could have hoped to get something of the order of Rs. 5 to 10 crore, which is the minimum you need to set up a lab working in quantum tech.

Was this part of the Rs 8,000 crore quantum national mission announced in the 2020 budget? 

No, this was a different programme where the first discussions started in 2017 and the money started coming in  2019-2020. Unfortunately, it was right in the middle of the pandemic, when you know, things were shut, and we couldn't spend the money. On top of that, as part of the pandemic-related things, the government made it very difficult to import things.

The National Mission on Quantum Technology and Applications (NMQTA) with a proposed budget of Rs 8,000 crore is expected to launch sometime soon. Probably, you know, in the next few months. We have been working on it.

Can you tell us more about the import policy?

It was introduced in the summer of 2020. This was not a science policy. This was a government procurement policy. You know, on the face of it, it seems like a good idea. They wanted to encourage Indian industry and push for self-reliance. On paper, it sounds great. And it might even work for certain types of government procurement.

They're basically saying that if you're supported by government funds, you have to prioritise Indian-manufactured goods. And the way they implemented it is by saying that if something costs more than Rs 5 lakh and you want to import it, you have to go through a new procedure and get a waiver before you can even float a global tender.

In the beginning, we were surprised, and we said, okay, maybe they just didn't think about scientific procurement. So, in a few months, they will fix it but it didn't happen. It took about six to seven months before some exception was given. But there is no blanket exception given for scientific procurement.

They reduced the number of signatures required by one and made some exceptions. Unfortunately, even there, because of a lack of clarity, every institute has interpreted the rules differently. And we are definitely facing a lot of problems still. In the best-case scenario, we are still wasting about three months just to get approval to import an item.

Does most scientific procurement happen from abroad?

The problem is to assume that there is an Indian option and we are actively choosing not to use it. That is not the case for 99 percent or even more. It is a well-known fact that for scientific high-tech equipment, there is no manufacturing happening in India… there are no companies which exist. But, we have to now prove it with paperwork to get the waiver. Even if we invest Rs 8,000 crore, it won’t work if these kinds of policies are not fixed. We are wasting a bulk of the time just doing this paperwork, and not actually doing research.

It might be more fruitful if you create opportunities in the nation for companies to come together and manufacture at least some of the things in India. But, when you have not had a manufacturing base in a certain technology for decades, how can you start doing that overnight? And in many cases, it might not even make economic sense for a company because it's a niche market.

Is it possible for India to build an indigenous quantum computer? 

In today's globalised world, nothing is going to be fully indigenous. For superconducting qubits, you can innovate the design of the chip and make it here, but the lithographic machines and dilution refrigerators have to come from outside.

As I said earlier, none of the available quantum computers are doing anything useful better than classical computers at present. When it happens in the future, and it will, you might be able to use cloud-based systems for many applications. On one hand, we can continue to be just a leader in the software layer. One could decide that we will develop the best quantum software in the world, but we will never develop any hardware.

The problem really comes when you start talking about strategic applications – whether the cloud company will provide you with the best systems they have access to. That's the fundamental question we have to ask ourselves.

We don’t have a quantum computer, but startups are providing a solution called quantum key distribution (QKD) for secure communications. How do we know it is really secure if they have not been tested against a quantum computing hack?

QKD fundamentally relies on the fact that if somebody tries to interfere, you will know. And the trick in any secret communication is that if two parties can exchange a key without anybody else knowing, then the information you transmit using encoding based on that key is secure. Quantum key distribution is a method where the laws of physics guarantee that the key you have exchanged has not been intercepted by anybody else. Because the moment somebody intercepts, you will know and you will discard that key. It cannot be broken by even a quantum computer. So you don't need to test it against a quantum computer.

A bureaucrat was saying that enterprises are concerned that they might be sold classical computing solutions in the garb of quantum tech. So the government is trying to bring a quantum certification programme. Would it have the bandwidth to do such accreditation?

You have to look at this slightly differently. There are things in quantum which are impossible in classical. So if whatever you're purchasing is actually carrying out that task, it cannot do it the classical way. And if there are applications where both classical and quantum can do that, then it is unclear why you would bother unless there is a cost advantage or a performance advantage for the quantum solution.

As a business or an enterprise that wants some problem solved, you will put the problem statement, and the quantum company will say this is the solution. So, why do you care what is inside?​ The bureaucrat was probably referring to quantum communication systems where you want some central body to verify that the system is really secure and meets certain standards. I believe the government is setting up such bodies for various quantum technologies.