At Sovetjheza Senior Secondary School in the rural village of Matshiding, Mpumalanga, Unathi Skosana remembers being fascinated by the physics teacher singing the Periodic Table Song in Ndebele, or conjuring up elephant toothpaste from a mixture of hydrogen peroxide, dish soap and a few drops of food colouring.
It is perhaps no wonder then, that when he was exposed to quantum mechanics in his third year at Stellenbosch University, that fascination with science kicked in again.
Today he is pursuing an MSc-degree in quantum computing, with his first research paper, written in collaboration with his study leader Prof Mark Tame, published in Nature Scientific Reports, one of the most authoritative scientific journals in the world of science. The article, a proof-of-concept demonstration of a quantum order-finding algorithm for factoring the integer 21, has already been downloaded 350 times and cited in two articles by researchers from Singapore, Malaysia, China, Spain and Germany.
Prof Tame, who holds the South African research chair in photonics in SU's Department of Physics, says he was delighted when Unathi decided to continue the work he started as a BSc Honours student on demonstrating small-scale quantum algorithms on IBM's quantum processors.
“For his MSc, Unathi decided to look at the prospect of realising Shor's algorithm on the IBM quantum processors for factoring the number 21. He is a very bright and promising researcher, and certainly does not shy away from tackling difficult challenges in quantum computing," he describes his student.
Shor's algorithm, developed by the American applied mathematician Peter Shor in 1994, is regarded as one of the crown jewels of quantum computing. Harnessing the laws of quantum mechanics, the algorithm introduced a completely new and efficient way of factoring numbers, which is considered to be a difficult problem for traditional computers. That is why the problem of prime factorisation lies at the heart of the RSA public-key encryption method, which is the basis of internet security as we know it today.
Unathi explains: “A key part of the 'quantum' improvement is because traditional computers process information that is stored in bits, which can take on values of '0' or '1', but never both. On the other hand, quantum computers store and process information in quantum bits or qubits, which operate and behave in a fundamentally different way. For instance, a qubit, just like a bit, can represent information as a '0' or '1', but in addition can also be in a combination of both '0' and '1' at the same time, called a superposition. By utilising the superposition feature, and other strange phenomena in the world of quantum mechanics such as entanglement, quantum computers will be better and faster than traditional computers at solving some computational problems."
Yet, despite several attempts over the past decades to realise Shor's algorithm for small numbers such as 21, the results have been noisy and not very conclusive.
In order to tackle this problem, Unathi first tried realizing Shor's algorithm in its original form on newly-built IBM quantum processors, and struggled with it for some time, getting very “noisy" results and nothing conclusive, as in previous attempts. In quantum computing, “noise" or “decoherence" happens when there are too many operations associated with an algorithm.
“Quantum technology is still in early development stages. This is primarily because of the inherent instability of qubits, which makes them very prone to errors," Unathi explains.
“In practice, the quantum nature of qubits is a doubled-edged sword, because it also makes them extremely sensitive to unintentional influences from their surrounding environment. The longer the computation goes on, the more this effect becomes pronounced which ultimately ruins the computation, giving unreliable results," he continues.
The breakthrough came when they played around with the idea of shortening Shor's algorithm for factoring 21 by replacing Toffoli gates – a universal logic operation used in traditional computing – with more compact logic gates and some clever maths.
“We were surprised at how good the results were, so we decided to write them up and submit a paper to Nature Scientific Reports," Prof Tame explains. The paper, titled “Demonstration of Shor's factoring algorithm for N=21 on IBM quantum processors" was published on 16 August 2021.
Prof Tame says he started experimenting with IBM's quantum processors in 2019. At the time, he wasn't that impressed. However, since then new processors with much better quality have been introduced on a regular basis, allowing him to consider their use by his postgraduate students. He submitted a proposal to the University of Witwatersrand for use of the IBM quantum processors under the banner of the African Research Universities' Alliance (ARUA) – a network of 16 of Africa's leading universities.
“Access to the IBM quantum processors provides an opportunity for students from South Africa and Africa to learn and develop skills in this emerging technology, and even play a leading role," he says.
He has recently incorporated the use of IBM quantum processors into his BSc Honours lectures on quantum information: “Students can run quantum computing tasks on real quantum processors and submit their results to me for checking. In future I am hoping to open up the course to non-physics students from computer and data science as well," he adds.
In the meantime, putting together the quantum hardware needed to do even small-scale quantum computing is expensive and time consuming.
And that is why Unathi is now tackling another challenging task: building a proof-of-concept small-scale quantum computer, based on the use of particles of light (called photons). The idea is to build something that will produce similar results to the IBM quantum processors, but in a way that is better suited to the use of photons.
“Imagine giving students access to our own quantum processor," says Prof Tame. “The theoretically-minded ones can immediately put their ideas to the test, while those interested in experiments involving quantum mechanics can come in and play around with the real thing."
Unathi seems to be more than up to the task when he mentions another major inspiration: the Greek mathematician and geographer, Eratosthenes, who, two thousand years ago, made the first measurement of the circumference of the Earth working only with the midday shadows cast during an equinox.
So who's to say it's impossible to build your own quantum processor in the Merensky-building here on SU's campus!
On the photo, from left to right, Prof Mark Tame and MSc-student Unathi Skosana from Stellenbosch University's Department of Physics. Photo credit: Stefan Els
Prof Mark Tame
SARChI research chair in Photonics, Department of Physics, Stellenbosch University
Office: +27 (0)21 808 3375