Late last year, the world of quantum computing achieved a significant milestone. A quantum processor developed by Google was able to solve a complex mathematical calculation in just three minutes and 20 seconds. In comparison, the most advanced supercomputer to date would have taken more than 10,000 years to solve it.
This significant achievement did not go unnoticed by Joe Fitzsimons. In fact, the former quantum computing professor had anticipated major strides such as this to happen. That had inspired him to leave research to set up Horizon Quantum Computing two years ago. Google’s achievement, he says, marked a shift in the “centre of gravity” of quantum computing towards industry from academia.
“When I set up the company in 2018, I felt we were getting to a point where there were going to be some advances quite soon,” Fitzsimons told The Edge Singapore in a recent interview. “We knew this was coming well ahead of time. I mean, everyone in the field knew what would happen relatively soon. So, it seemed like the right time to move into a start-up if I was ever going to do it — to be ahead of that than after that.”
Horizon aims to bridge the gap between quantum computers and conventional software development. The start-up is developing tools to automatically accelerate programs written for conventional computers, and adapt them for quantum processors. These tools will also make life easier for experienced quantum software developers, by allowing them to construct complex programs with fewer lines of code and by automatically optimising their programs for real processors.
Fitzsimons says these tools could still be developed had he remained in research. But he is aware that a wider set of skills is required to build useful technology for real world applications. “So you need to be able to hire not only the people that are experts in quantum computing, but software engineers, web developers, designers — different kinds of people with different skill sets, who can come together to build a single thing. That’s harder to achieve in academia,” he says.
Funding is also another important aspect to develop such tools. The private sector has deeper pockets compared to that of academia. In June, Horizon announced that it has increased its seed financing to $4.5 million, thanks to a seed-plus round led by Sequoia Capital India. Previous investors such as SGInnovate, Abies Ventures, DCVC, Qubit Protocol, Summer Capital and Posa CV also participated in that round. “I think a lot of investors have seen the value or potential value in Horizon,” says Fitzsimons.
Quantum software hurdle
Today, advancements in technology have brought quantum computers closer to reality. But there is still some way to go before the next generation of computers get rolled out. Hardware continues to be one of the main challenges to building a quantum computer, according to Fitzsimons. It needs to be sufficiently sophisticated to be able to support useful computations and economically viable to justify the cost, he says. The other main challenge is software. “It’s no good just having the hardware if you don’t know what to do with the hardware. So that’s where we put our focus on,” he explains.
For the uninitiated, quantum computing does not process information in the conventional binary way — that is, using bits that represent either 1 or 0. Instead, it does so via a quantum bit or qubit, which stores a combination of 1 and 0. Existing programs are not able to understand and communicate with a quantum processor.
This is because an existing compiler — a software that translates source code written in a high-level language into a set of machine-language instructions — is only able to do so for conventional processors.
To solve this problem, programmers may choose to rely on libraries of pre-programmed algorithms or require programs to be written using gate level languages, says Fitzsimons. However, he reckons that a better way is to build a compiler that can understand and communicate with a quantum processor.
Such an approach — which is taken by Horizon — alleviates the need for any quantum knowledge. “We’re trying to do it in a way that the person writing the code does not need to understand how the hardware works out. So that they just need to think about it as if they were developing software for a conventional computer,” says Fitzsimons.
This approach is also hardware agnostic, though the industries that benefit the most are the ones that require high-performance computing to solve complex problems, says Fitzsimons. These include aerospace and automotive companies that are required to simulate airflow over an aircraft wing or automobile.
“So, the hope is quantum computing will be able to significantly reduce the resources you need to do some of these calculations. And by reducing the resources you need, [hopefully] they’ll also open up new things that you couldn’t do before, because they would have been too hard to compute, even for the most powerful computers available currently,” he says.
Thus far, Horizon has developed several prototypes “of a number of different stages” of the compiler, says Fitzsimons. The company has successfully demonstrated the proof of concept of these prototypes earlier this year. It is now in the process of integrating these prototypes into a single product, he adds.
Literary inspiration
Like many of Fitzsimons’ peers, getting into the field of quantum computing was sort of a natural progression. He had always loved mathematics, physics and computer science when he was in secondary school.
But what really pushed Fitzsimons into the world of quantum computing was when he read a book called The Code Book: The Science of Secrecy from Ancient Egypt to Quantum Cryptography. It was written by Simon Singh, a well-known science author, theoretical and particle physicist in the UK.The book talked largely about the history of cryptography and how quantum computers could be used to break cryptography. That piqued his interest in quantum computing.
And so when Fitzsimons had the opportunity in college to take a course in quantum computing, he jumped at the chance. “And from then on, I decided to do a PhD and continue on in that area. I’ve always wanted to work on forefront technologies,” he says. Fitzsimons holds a doctorate from the University of Oxford on quantum computing architectures.
“Quantum computing was really the most obvious technology that stood out as a combination of my interests. It is also a way of maybe reliving some of the excitement of the early days of computing when computers were first being built. And I thought that will be an immensely exciting thing to be part of,” he adds.
Prior to founding Horizon, Fitzsimons led the Quantum Information and Theory group at Singapore University of Technology and Design (SUTD), where he was a tenured associate professor. He was also a principal investigator at the Centre for Quantum Technologies (CQT), hosted by the National University of Singapore.
At SUTD, Fitzsimons met his co-founder and chief science officer Tan Si-Hui, who holds a Bachelor of Science with honours in physics from Caltech and PhD in Physics from the Massachusetts Institute of Technology. She had previously worked at CQT and Agency for Science, Technology and Research’s (A*STAR) Data Storage Institute.
Fitzsimons says the big dream for Horizon is to widen the applications of quantum computing far and wide, and not be merely confined within computer laboratories. “We see quantum computing as a general-purpose computing technology. A lot of others in the field see it as something that will be used as a co-processor, as a kind of special purpose accelerator that will be used to speed up a small number of computations,” he says.
“But for me, it’s hard to imagine a world where we don’t do computation using quantum computing instead of conventional computing. It’s really the next wave of computing, or maybe the wave after next. But it’s really a transformation in computing,” he adds.
