For more details refer to the documentation.
- Security patch: require twisted 20.3 or higher
- Fixed bug when releasing qubits.
- Major refactoring of
cqc.pythonLibwhich splits the code into multiple files and improved extendability by the use of abstract classes etc.
- Allow for numpy version up to 2.0.0
- Pending messages are now flushed when a connection is closed.
- CQC now supports logic. That is one can send a batch of CQC-instructions to the backend which can have conditional logic based on measurement results.
Note, that this is different from having logic in the application in the Python library since this requires communcation back and fourth from the backend to the application.
The Python library is also updated to be able to construct these instructions.
For examople:
- to apply instructions a certain number of times you can now do:
from cqc.pythonLib import CQCConnection, qubit, CQCMix with CQCConnection('Alice') as node: # qubit is created beforehand qbit = qubit(node) # Start of the CQCMix with CQCMix(node) as pgrm: qbit.X() # Start of the Factory # Apply H three times with pgrm.loop(times=3): qbit.H() # Y gate which is not part of # the loop (i.e. Factory) above qbit.Y()
- or to perform certain instructions based on a measurement outcome you can do:
from cqc.pythonLib import CQCConnection, qubit, CQCMix with CQCConnection('Alice') as node: # qubits are created beforehand qbit1 = qubit(node) qbit2 = qubit(node) # Start of the CQCMix with CQCMix(node) as pgrm: result = qbit1.measure() # if measurement yielded 1, apply X with pgrm.cqc_if(result == 1): qbit2.X() # else, apply H with pgrm.cqc_else(): qbit2.H()
- to apply instructions a certain number of times you can now do:
- If a CQCConnection does not manage to connect to the cqc server, the used app ID is popped from the list of used ones, to be reused again.
- Fixed bug that mixes up return messages for different application IDs
- Updated coinflip example and added CI.