The innovation race on Quantum Computers – where are we today?

I had the privilege to be a speaker and panelist at the TCS Quantum Symposium which was held in IIT Bombay on April 7, 2019. The event was an opportunity to have an open discussion with the most intelligent minds exploring the practicality of Quantum Computers today and its potential breakthrough in applications.

I was fascinated to see the problems that researchers are trying to solve and the results they were able to obtain for extremely small problems to potentially evaluate where we are from the practical point of view.

Quantum computing could be the most discussed and investigated innovation in recent times. Every large IT service provider and organization is keen to take a potential lead in this technology. But delivering practical results from Quantum still seems some distance away.

This was one of the most discussed and debated topic in this symposium, from all IT solution providers like TCS, IBM, 1QBit and of course Fujitsu.

In the realm of software, quantum algorithms have matured to a reasonable extent. These now show improved results for various problems being explored.  However, the underlying hardware has not been able to provide the best possible solution.

Some of the most complex challenges being examined include quantum chemistry, computational chemistry and various other combinatorial optimization problems such as traffic route optimization, portfolio risk optimization, robot positioning, and so on.

For problems such as these, existing quantum computers fail to give a better result than classical computing system as the problem scales to a real-world data set. In quantum annealers, the problem is called ‘chain break’ – essentially the problem which is embedded breaks as the scale increases resulting in non-optimal solutions or errors.

There is often a misconception about physical and logical qubits. The quantum algorithm requirements are defined by the logical qubits required which may vary from 20qubits to 50qubits, and also to even 1000qubits, depending on the problem size and complexity.

However, when you state there is a quantum computer that supports 50qubits or 100qubits – these are essentially physical qubits and don’t translate directly to logical qubits due to 2 factors- “noise” and the fact that qubits are not universally entangled, which brings in the need to have the requirement for error correction.

The translation from physical qubits to logical qubits basically depends on the quality of the qubits or circuits and the requirement can be anywhere between 100 to 10000 physical qubits per logical bit. So, potentially solving a real-world scale problem being discussed could need over a 1,000,000 physical qubit scale.

This was an accepted and understood fact as the session moved forward with some discussion around the results obtained on true quantum computers.

It was really encouraging for me to see how all the innovators from the IT industry who are developing quantum computers were open in acknowledging this reality and discussing how could this be potentially solved in the future.

I was also delighted to see the eagerness in understanding how Fujitsu’s digital circuit architecture was able to deliver and solve some of these combinatorial optimization problems today.

Our Digital Annealer is helping to address challenges such as portfolio risk optimization, molecular similarity, vehicle route optimization and robot positioning optimization. This is providing a game-changing advantage to organizations and also helps prepare for the Quantum computing era when it is ready.

I did see a lot of perplexed faces when explaining the reality of something that seems impossible. If you're amongst them, this video should help you understand ‘Quantum-Inspired’ computing technology.

Interested in finding out more? Visit our website for all the details.