So the big news today in quantum computing came from IBM — here is the MIT Technology Review’s announcement, plus Tech Crunch’s. Basically IBM has prototyped a 50-qubit machine, and they are allowing the public to use a new 20-bit machine via their QISKit library. In the past, they allowed the public to use 5-qubits and researchers to register for 17-qubit access, but now everyone will have access to 20! It will be interesting to see what additional research folks can carry out now…I recall reading that there have been ~30 papers published using the IBM system, so researchers should be able to do every more things now.
I have also been spending some time reading through the tutorial Jupyter notebooks for QISKit … wow. The library still seems to assume a strong grasp of quantum mechanics. It allows you to do things like set gates and levels in the quantum circuits, which is pretty cool. But you have to understand how to calculate the energy equation for your problem in the first place (and then, how the combination of gates defines your energy function — this, I think can be codified). I would love to see either or both of:
- More tutorials on how to construct quantum algorithms from real-world problems. Ising what? Do I need a Ph.D. for this?
- Libraries that abstract away the quantum weeds. So I don’t need to manipulate and set the individual gates, the possible combinations should be abstracted away by some higher-level command. I feel like this is simpler than the first bullet, honestly.