The device is unremarkable to look at: a slab of patterned metal on a cooled substrate, photographed in a presentation case. What it did last month was run a sampling problem in minutes that would take a conventional machine many years.
The claim of "quantum advantage" is now familiar, and so are the caveats. The benchmark has no practical use, error rates remain high, and classical algorithms have a habit of catching up once a target is published.
What is new is the engineering: a larger number of qubits held stable for longer, and control electronics that scale. The road to a useful machine is still long, but this year it looks like a road rather than a hope.
Comments
Sign in to join the conversation.
No comments yet — be the first.