Quantum computers will solve problems that are beyond the reach of conventional computers. The only hitch is that such computers are yet to be built. Professor Zoltán Zimborás does not let that slow him down: he is busy developing software for future computers.
A lot of expectations are heaped on quantum computers, and not for nothing. They are something entirely different from all computers seen to date.
"All computers developed so far are variants of a Turing machine, simple tally marking," says Professor of Quantum Algorithms of the University of Helsinki.
"Computers complete one calculation at a time, store the results and combine them. Ultimately, their only advantage over simple tally marks is speed: supercomputers can complete more than two quintillion calculations per second, and even ordinary desktop computers can manage a trillion. In principle, however, the same calculations can be completed more slowly by hand too."
Quantum computers are in a different league. They base their calculations on particle-level phenomena that go against human understanding, such as entanglement and superpositions. With the help of their computing units that are in quantum state, known as qubits, they solve calculations not consecutively, but simultaneously.
This cannot be done with tally marks.
Shor's algorithm changed everything
The idea of quantum computers is not new. It was first introduced by physicist Richard Feynman in a speech given in 1981.
"Feynman was frustrated by the slow and ineffective pace of modelling quantum physics phenomena on conventional computers. He therefore proposed his idea for a quantum computer that would compute such phenomena by its own rules."
Feynman's initiative long remained purely theoretical. In 1994, the American mathematician Peter Shor nevertheless surprised the world by unveiling the first working quantum algorithm. Using this algorithm, coined after Shor, a quantum computer effortlessly breaks down numbers into their prime factors.
This way, it is able to very quickly decrypt most current encryptions that are based on the products of prime numbers.
Or it would be, if the algorithm could actually be used. The quantum computers so far designed have been too small and unstable to run the algorithm.
However, this does not preclude the development of new quantum algorithms even before machines are built that can run them. This is what Zimborás and his research group are working on.
"This way, we can anticipate the moment when quantum computers actually become useful."
Among other things, the group are engineering an algorithm that would help model complex molecules at the atomic level.
"Calculating quantum world interactions is very cumbersome with conventional computers. It would take millions of kilobytes to calculate the interactions of just a few particles," Zimborás says.
"An error-free quantum computer could theoretically complete such calculations with considerably less effort using its qubits, only a few of which would be needed for each electron in the system."
A ten-year wait
Error-free quantum computing is a tall order. All quantum computers so far developed are sensitive to disruptions and make mostly incorrect calculations, which is why much of their power is spent on screening for correct results from the noise.
Here too, Zimborás and his group are developing solutions.
"We have been developing an algorithm to obtain reliable results. It would connect 91 qubits, as it were, into one reliable qubit."
For today's quantum computers, even this is a difficult task, as the largest functioning quantum computer comprises roughly 100 qubits.
What, then, could quantum computers be used for besides deciphering passwords? Big expectations have been set for tasks involving enormous amounts of calculations, such as route optimisation or the development of novel drug molecules. Zimborá s believes that the first applications for quantum computers will be associated with modelling phenomena in the quantum world as envisioned by Feynman.
However, practical applications are still to come. While the most optimistic people are talking about a few years, Zimborás himself believes that a really useful and well-functioning quantum computer will be achieved in roughly ten years.
"You can consider that a short or long time. In any case, it's a near future that we can prepare for by developing algorithms now."
