Cloud Hosting

Phil Morle, a partner at the CSIROs Main Sequence Ventures fund who specialises in this kind of investment, says what happened to SQC is typical of whats happening across the sector: investments are chunked down into smaller rounds with shorter, more demanding timeframes.

Theres a sort of meta dynamic thats happening across investments ... where theres a lot more caution, and a lot more expectation of evidence for early revenue. Its something which deep-tech companies need to pay very strict attention to, he says.

Where a couple of years ago we may have collectively been able to have a longer range plan that drove the technology towards commercial outcomes a bit later, thats not on the table today.

That shift in investor expectations has caused a re-alignment at SQC, both in terms of its rate of growth had it raised $150 million, there would be more ambition to expand the company quicker, Professor Simmons says and in terms of the technology its focusing on.

The company now talks about quantum simulation, as opposed to quantum computation, much more than it did in the past.

I think everyone ultimately wants the programmable computer, but on the way to getting there everyone recognises that nearer-term commercial products are more likely to come out in the simulation space, Professor Simmons told AFR Weekend.

Luckily, it turns out that SQCs plan to build a quantum computer, first proposed by the American physicist Bruce Kane when he was visiting UNSW as a student in 1998, is exceptionally well suited to quantum simulation, too, she adds.

Kanes ambitious plan was to place phosphorous atoms on to a silicon chip, aligning them with such precision that they would interact with each other to form a quantum computer.

The plan should SQC manage to pull it off would mean a quantum computer with vastly more qubits than any other. Each of SQCs qubits is only a single atom in size, where competitors such as Google and Microsoft are building qubits out of superconducting circuits that, while tiny, are measured in tenths of millimetres, rather than in tenths of nanometres.

Some 25 years on, SQC has developed that atomic alignment technology to the point where the company can now churn out experimental new versions of its quantum computer every week, Professor Simmons says, inching its way towards building an error-corrected quantum computer by 2033.

Error correction, which for the most part involves entangling tens or hundreds of physical qubits together to form a single logical qubit, is how the industry deals with the perturbation problem.

A stray cosmic ray might disrupt one physical qubit, but shouldnt be enough to throw off an entire logical qubit, the theory goes.

SQCs most recently revised road map has it delivering a single, error-corrected qubit made up of 100 physical qubits by 2028. By 2033, it is building an error-corrected computer thats powerful enough to be useful to a broad audience of users across multiple use cases.

(The roadmap has shifted over the years. Professor Simmons had initially hoped to build a prototype 10-qubit quantum computer by 2020, but later revised that date to 2023. While SQC is still on track to produce such a prototype this year, its the error-corrected milestone thats now the important one, she says.)

The rapid iteration that would get them to the error-corrected quantum computer happens to work well for quantum simulation, too. Etching a model of a molecule on to silicon just to perform a single calculation isnt such a grind when you can churn out new models quickly.

Our approach does translate unusually (or exceptionally) well to simulation. We manufacture devices quickly within a week, she says.

Even so, revenue from the easier quantum simulation technology is still three to five years away, says Simmons.

Whether thats fast enough for investors, only time will tell. But, with some quantum computing start-ups already putting revenues on the books, the pressure on SQC to produce revenues has only increased in the past 12 months, making its task more difficult than ever.

Q-CTRL, a quantum computing startup at Sydney University that makes error correction software for quantum systems including quantum computers, announced sales of $15 million in 2022, and was cashflow positive for the first half of 2023, its founder Michael Biercuk says.

Almost all the $US54 million ($84 million) Q-CTRL raised this year in its Series B fundraising is still sitting in the bank, he adds.

Investors are no longer saying Gosh, one day when quantum computing is going to be big, but who knows when? That is no longer the conversation, says Main Sequences Morle, whose fund invested in Q-CTRL

Quantum computing is absolutely commercialising now, and investors are starting to look at it through that lens.

Go here to see the original:
Building a quantum computer was always hard. It just got harder. - The Australian Financial Review