Thread

>Somewhere in that gap between 24 qbits held for a few microseconds and ~2000 held for hours to do shor on one Bitcoin key, the ceiling is lurking. Ok so now we're getting somewhere. Right then, Atom held around 28 logical qubits and ran Berns-Vaz algo w demonstrable error correction, under a second. That was 2 years ago, a lot has happened since. And shor's we need ~2000 and hours. I'd argue that for shor's it's more likley around 1,700, we've already optimised it down by half, and there's more optimisation in the tank. pure math by the way. so by offloading the most complicated steps (modular arithmetic and fraction conversion) to highly efficient classical methods, the quantum part gets streamlined. You're saying somewhere between 28 and 1 second and (if optimisation holds)1,700 and hours is some arbitrary limit of the universe. Atom will ship commercial machines 48 next year. Means them + Microsoft, machiens at 48 logical, capable of running deep, sustained computations for minutes, hours, or potentially days in 2026. (this is all proven out in their lab, to prove me right that's just waiting 6 months to ship. (it is right though) So 48, and running for hours. Where does it stop? At 96 and days? If not at 96 and days then at 192 and weeks? If you can't say where the limit is, or even give a range, based on some physical properties, then you are basing the limit on nothing. You have to give a number to show you've got some physical basis for your impossible statement.