Support concept of static qubits

[denialhaag]

In our QC and QCO dialects, we have the concept of static qubits, allocated via qc.static and qco.static, respectively. For more information, see here: https://mqt.readthedocs.io/projects/core/en/latest/mlir/QCO.html#qco-static-qco-staticop

Does it make sense to support this concept also in jeff? The concept would be useful for abstraction levels closer to the hardware.

[burgholzer]

Maybe just for context:
What Daniel is referring to here is the concept of physical qubits in OpenQASM 3 (https://openqasm.com/language/types.html#physical-qubits; denoted by $<index>) or the concept of static qubits used in the base profile of QIR (https://github.com/qir-alliance/qir-spec/blob/main/specification/profiles/Base_Profile.md#qubit-and-result-usage; denoted by ptr inttoptr (i64 <index> to ptr)).

While these are, in principle, low-level concepts, it may make sense to support them as part of the Jeff exchange format.

Both in Qiskit, as well as in our mqt-cc, routing/mapping methods transform circuits using dynamically allocated (virtual/program) qubits to circuits using statically allocated (physical/hardware) qubits. If circuits are to passed back to another tool after routing, the exchange format would have to support the respective concept.

[dime10]

Interesting idea! I wouldn't be opposed adding these to jeff 👌