Skip to content

Conversation

@TysonRayJones
Copy link
Member

No description provided.

although postMultiplyFullStateDiagMatr may require a bespoke backend function
which required updating the backend functions since FullStateDiagMatr makes use of an optimisation that precludes twice-enumerating the density matrix
as a template for other doc
which required that we undo the transpose effected upon the operation which results from operating upon the bra-space of the vectored density matrix
which required undoing the transpose upon the operation which exists from operating upon the bra-qubits of the vectorised density matrix. This required...
- giving the backend any-sized dense matrix function an additional ApplyTransp template (and updating all the associated function instantiation/dispatching - happily simplifying it)
- we hijacked the cuQuantum backend (which can effect the adjoint) to effect the transpose by pre-conjugating
- transposing 1 and 2 qubit dense matrices in localiser.cpp
@TysonRayJones
Copy link
Member Author

Paid tests passed, except for a now-patched validation issue (use of a density matrix over statevector un-triggered distributed applyCompMatr validation)

@TysonRayJones TysonRayJones merged commit 738ca89 into devel Jun 30, 2025
130 checks passed
@TysonRayJones TysonRayJones deleted the post-multiply branch June 30, 2025 18:04
@TysonRayJones TysonRayJones mentioned this pull request Oct 13, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants