Researchers have developed a new protocol that enables quantum unitary channel tomography with parallel queries as efficiently as with sequential queries. Traditionally, sequential strategies were considered more powerful because they allow the learning strategy to be adapted as the channel estimate is refined. Parallel queries, on the other hand, do not allow such adjustments, which suggested an inherent limitation in their efficiency.

The new protocol achieves learning an unknown d-dimensional unitary channel to within ε in diamond distance using O(d²/ε) parallel queries. This result matches the previously established theoretical lower bound, thus closing the efficiency gap that existed between parallel and sequential quantum tomography strategies. This advance is significant because it demonstrates that, contrary to previous intuition, parallel queries can be as efficient as sequential ones for this specific task.

Applications of this protocol include optimizing query complexity for quantum channel tomography in the boundary regime and improving the query efficiency of the best-known protocol for fermionic linear optics tomography. Achieving optimal efficiency with parallel queries opens new avenues for the development of quantum technologies that require fast and precise characterization of quantum processes.