Q-CTRL Shows Significant Improvements in Quantum Algorithmic Success with Its Quantum Control Sof...

One of the biggest limiters in quantum computing today is the inherent error rate in qubit gate operations that will cause the processor to make incorrect calculations. Currently, this problem gets worse as the number of qubits increases and the number of gate levels in the algorithm increases. This a major focus area for improvement for quantum hardware developers as the create their next generation systems. Although the long-term solution is believed to be quantum error correction where multiple physical qubits are ganged together to create a logical qubit with a much lower error rate, in the near term researchers are developing various error mitigation techniques to lower the error rates by other means including compiling techniques and more sophisticated control of the pulses that implement the gate operations in the qubits. Q-CTRL's Boulder Opal program is one such program that can help mitigate errors. It allows a user to characterize quantum parameters, design optimized noise-robust controls, automate hardware tune-up by leveraging AI, and benchmark and simulate performance.

Q-CTRL has announced they have achieved significant improvements in quantum algorithm success with the quantum processor benchmarking suite released last year by the Quantum Economic Development Consortium (QED-C) in testing that used its Boulder Opal software. The amount of the benefits could vary significantly depending upon the specific algorithm and processor used, but Q-CTRL did report achieving improvements in success probability of up to 9000X in a test that used the Bernstein-Vazirani algorithm with IBM's 16 qubit Guadalupe machine. Although other tests using different benchmarks or other processors showed lower levels of improvements, Q-CTRL's testing showed a consistent pattern of improvement over multiple runs on different days, different algorithms, and different processors. The company ran these tests on circuits with a mixture of different number of qubits and also saw an increasing advantage in using their software as the number of qubits increased.

Q-CTRL will be describing details of these benchmarking tests along with information about their software and performance improvements in six technical talks at the upcoming APS March Meeting on March 14-16, 2022. Until then, you can read a press release here from the company that describes the highlights of the improvements there were able to achieve.

March 8, 2022