Saturday, October 05, 2024

Op-Ed: The strange business of building the world’s first operational quantum computer


ByPaul Wallis
DIGITAL JOURNAL
October 5, 2024

IBM's quantum computer, London. — Image: © Tim Sandle

Australia, after years of research and development, is finally getting a real live solid quantum computer. The science is good. The business side is attracting a lot of dubious comments. This is far more than skin in the game. It’s a functional form of future computing science on the grand scale.

The billion-dollar investment will be located in Brisbane. It’s supposed to be a world first and is scheduled for completion in 2027.

It will be created by PsiQuantum, an American company.

Doubts include:

The practical functionality of the computer. What can we use it for? Never mind the hype. Practical applications are what this is all about.

The cost, as might be expected. Contract blowouts are not popular.

PsiQuantum says they’re a for-profit company, so that’s already another separate cost vs value issue.

Operational costs, including cryogenic cooling, components, and maintenance. One operation per year doesn’t yield a good ROI.

Energy costs are not publicly specified.

Computer specifications provided by PsiQuantum are considered to be vague or inadequate. These are opinions, but they are relevant.

Who gets access to PsiQuantum services is unclear.

Public funding issues.

If this sounds like a recipe for disaster, it’s not. Not quite. At this point, the doubts are reasonable but lack their own specifics. This is pretty ordinary, if not encouraging, in the tech sector. We have a billion-dollar controversy about something that doesn’t exist yet except as a basic idea.

This is fairly normal in Australia. We try to keep up with the vacuous indecision of other countries, and we’re getting pretty good at it.

All of the above obscures one major unavoidable issue. Quantum computing is more typically done in a lab. It’s nothing like mainstream science, let alone business tech. A baby quantum computing science needs to learn to walk.

We need at the absolute minimum:

A rock-solid fully costed and verified contract for construction, obligatory services, and support.

A functional map for testing and evaluation. This is proof of it doing what we need it to do.

System quality control for the duration of operations. We can’t simply take it for granted that any operational functions will be easy or simple.

Quantum computing has great potential. This could be an extremely valuable long-term asset. Coupled with AI, it could be far beyond any previous computing tech, not just faster.

So much for the sales pitch. I’d also like to point out that any disincentive to investment in quantum computing would be a major own goal for this tech, It’s not just PsiQuantum who’ll be affected if it screws up.

This project will be the face of quantum computing in the business, tech, and investment sectors. They’re still trying to get their heads around the basics of AI. They’re not necessarily inclined to invest in added tech that’s still in the bassinet.

PsiQuantum needs to look good and work well for the sake of future science.

No comments:

Post a Comment