Quantum start-up Q-Ctrl “unmixing the soup” of qubit decoherence
- 03 November, 2017 12:24
Photo: Michael Biercuk
Before quantum computers can be truly realised, there is a major hurdle scientists must overcome.
Quantum systems are highly susceptible to decoherence. The fundamental building blocks of a future quantum computer – quantum bits, or qubits – degrade rapidly, their states randomised by interference from the environment.
The result is the qubits have a shortened useful lifespan, seriously limiting the viability of quantum technologies.
Now, an Australian start-up – Q-Ctrl, led by University of Sydney Professor Michael Biercuk – is working to address the issue with a suite of controls that can stabilise the fragile systems, and “effectively turn back the clock” on decoherence.
“So all the randomisation that occurs, unwinds; it’s like unmixing the soup,” Biercuk explains. “You have to obey certain rules, you can’t just do it willy-nilly. But we know a set of techniques that allow us to unwind this phenomenon of decoherence in certain circumstances.”
Q-Ctrl – the first spin-off company of the Australia Research Council Centre of Excellence for Engineered Quantum Systems (EQuS) – launched this week with a slice of the $200 million CSIRO Innovation Fund managed by Main Sequence Ventures, plus further funds from an overseas venture capital firm.
“We aim to become the trusted provider of quantum control solutions for all quantum technologies,” Biercuk said.
Biercuk points to a classical computing analogy: the ‘memory refresh’ of DRAM in conventional computers.
“There was a problem being attacked in that the charge on the capacitor would leak into the substrate. And when it would leak it would randomise the information. They faced a challenge in that engineering community. They can either try to come up with a hardware solution, a totally different memory architecture [like SRAM] or say we're going to use a control solution,” Biercuk explains.
DRAM is refreshed – given a new charge – a thousand times a second, allowing the memory to be stabilised.
“It’s an amazing thing. You don’t need to do extensive, complicated hardware engineering. With the right control you make the memory live longer. We're doing similar things now obeying the rules of quantum physics,” says Biercuk, who spoke to Computerworld Australia on the sidelines of the the Gartner Symposium/ITxpo on the Gold Coast.
Q-Ctrl’s products apply controls one layer up in the software stack, at the virtual layer – in effect firmware for quantum computers.
“The firmware tools Q-Ctrl is building have had their performance validated in the lab and show orders of magnitude improvement in reducing qubit errors without the need for changing the underlying hardware,” Biercuk says.
The firmware will complement the algorithmic quantum error correction which takes place higher up in the stack. At present, running the correction algorithms is incredibly resource intensive, but combined with Q-Ctrl’s controls, become for effective and efficient.
“You need to do less, you need fewer resources, it makes the whole thing better. It’s in combination, it’s like a drug cocktail,” Biercuk adds.
Q-Ctrl is now seeking to hire back-end and user interface software developers (“all the skillsets that are not endemic to an academic research team,” Biercuk says) to produce more useable products. The team will grow to around 10 people and initially be based within the University of Sydney.
“We will be building products initially for the broader research community to try and give access to this black art our team specialises in,” Biercuk says.
“The science that we do is published, everybody has access to it, but it remains challenging to deploy these very technical solutions for organisations or people who are not specialist in this. In order to increase the uptake we know we need to have better software offerings to make it accessible.”
The tools are effective whatever the approach to creating qubits, with the potential to advance quantum systems “from the terrible things we have in the lab into things we need for computing,” Biercuk says.
IBM, one of the major players in quantum computing development, has already expressed interest in the company’s solutions.
Beyond the immediate demand from those involved in developing quantum computers, Biercuk sees future applications in quantum sensors, clocks and other hardware.
Avoiding a bloodbath
Q-Ctrl is the latest of a number of quantum computing companies to receive funding as excitement around the emerging industry grows.
In August, Australia's first quantum computing hardware company Silicon Quantum Computing launched with investment from USNW, Commonwealth Bank of Australia, Telstra, the Federal Government and the New South Wales Government.
A Morgan Stanley report last month said the quantum economy was set to double to $10 billion in the next decade.
Biercuk, although welcoming the heightened interest and funding in the field, urged patience around progress.
"It's extremely exciting to see this quantum economy building. My scientist hat says we need to manage expectations, because if we get unsophisticated investors who see the excitement and the hype, in a year when there’s no quantum computer, people are going to be upset and feel burned,” he says.
Gartner places quantum computing at the ‘innovation trigger’ stage of its emerging technologies hype cycle, the step before the ‘peak of inflated expectations’ which precedes the ‘trough of disillusionment’. It predicts mainstream adoption is more than a decade away.
“This is deep tech, this is a really hard, decadal-scale problem…If the field collapses because investors think that were going to have a Shor's factoring algorithm system in two years and it doesn’t happen it’s going to be a bloodbath,” Biercuk says.
“We need to build the right collection of investors. See the opportunity and excitement for what it is and see through the hype. Build the right mindset about what we’re trying to do and how epic it is.”
The author attended Symposium/ITxpo as a guest of Gartner