[dangirsh]

Note that this is now a reality: architectures very similar to slide 60 have been built and made cloud-accessible.

The note about "qubit division multiplexing" has also been built: some platforms provide access to a connected component of qubits instead of the whole chip. This is helpful for maximizing the availability of the hardware, since it's still extremely scarce.

[mikewave]

At D-Wave, our cloud-accessible QPUs are entirely available to one user at a time, in the form of "problems" that are submitted to a "solver" stack to load onto the QPU and sample in order to get an answer. This ends up being much like having a single-threaded CPU and a traditional OS handling multitasking; we have a scheduler higher up the stack responsible for determining which user problem to run next. The result is realtime performance for users while ensuring that the QPU sits idle as little as possible.

For the most part, there aren't enough qubits yet even on our 5000+ qubit machines that anyone would want to share the QPU with someone else during an execution cycle. In fact, this is the reason why we have a Hybrid Solver that combines classical and quantum technologies - people want to work on much, much bigger problems than fit onto the working graph of the QPU.