A new quantum computer in Denmark!

Alternative clickbait topic:

"How often are new quantum computers built?"

Quantum computers are wildly expensive, so you rent your computer time at the existing centres rather than getting one computer of your own. And now, there might be one more to share – might, because I am not certain about their policy regarding sharing. Magna is the new quantum computer that you can read about here (on the Danish Quantum Community website – I love the name). The expectations are that it could be ready in 2026, but possibly also in 2027. I shall ask if I can come to visit.

A person I met who has been working in the quantum computer domain is rather pessimismistic about it overall – that is why s/he left. To begin with, no one knows how quantum computers work exactly; if anyone claims they do, they are lying (my source emphasized this). Secondly, these computers are so powerful, that despite their massive calculation errors (more about that in the next newsletter), only a tiny fraction of their capacity is needed to calculate what we now expect to be "advanced molecule structures". That is how powerful they are, and this does not leave much else for us to compute, my source says. A quantum computer is, according to my source, a “fun physics thought experiment” with no real use.

I want to be realistic – but I am a bit reluctant to the claim that there is no future use. The beauty of research and technology development is that we will find new areas in life where our inventions and findings will be of use. We could not have predicted listening to music on CDs two hundred years prior – but we can predict that things can change and evolve in unprecedented ways. So, maybe additional applications will come with time, and we can use the knowledge we gain from the project in other areas meanwhile? I mean, the babies born today – the jobs they will have do not even exist yet.

I also asked a representative at one of the computer centres in the Nordics about the memory storage of a quantum computer. Is it long, short, how short?

Turns out, its memory is very short – and not transferrable. I had actually assumed that the quantum computers learn from their previous calculations, but it seems to not be the case at all because of the basics: the no-cloning theorem, which is about the inability to copy a quantum position without simultaneously destroying it. The quant processor is right now its own memory, starting over and over at each calculation.