Quantum Circuits is excited to report a key innovation in its development path toward scalable quantum error correction. With our dual-rail cavity qubit design, we have succeeded in creating both a high-speed and world-class fidelity entangling gate for the first time in the quantum industry.
This performance is critical for practical commercially useful systems. Speed is required to successfully support deep quantum circuits executing complex algorithms, while high fidelity is required for building scalable fault-tolerant quantum systems.
A universal quantum computer, like ours, requires a gate set that includes entangling operations. This is one of the amazing (and non-intuitive) cornerstones of quantum computing. Getting them right with high reliability is essential, because physical qubits on their own are too fragile to get us to quantum advantage – we need error correction.
In our latest arXiv paper, we demonstrate a novel two-qubit entangling operation that realizes a high fidelity Controlled-Phase (CZ) gate – closely related to the CNOT – with world-class performance metrics.
Our team has designed a new pulse-level protocol, known as the “Swap-Wait-Swap” (SWS) gate, which leverages the unique hardware properties of the Dual-Rail Cavity Qubit (DRQ). The SWS gate is a simple, yet clever pulse sequence that efficiently entangles the photons stored within the two DRQs. The SWS protocol results in a so- called CZ gate between any two adjacent DRQs, that is both fast (~500ns execution speed) and high-fidelity (exceeding 99.9% after erasure detection).
The higher the fidelity supported by the gates, the easier quantum error correction becomes. In fact, error correction approaches typically have what’s called a threshold – the minimum fidelities for these gates before error correction actually yields benefits. With our CZ at greater than 99.9%, matching or exceeding what most modalities are able to achieve today, we’re blowing past the threshold for the entangling gate to build scalable fault-tolerant quantum systems.
Now, if we couple that high gate fidelity with the speed of the CZ – 500ns – we also get a quantum system with superior performance supporting practical applications. Superconducting systems, such as ours, are perfect candidates for scalable quantum computers because of how fast they are. You want to have something that can run a large, commercially valuable application in a reasonable amount of time – in a minute, an hour or even a day…not months or years. In addition, you also want to do so with high fidelity – that’s what the CZ we report on today is giving us, the unique combination of both. And that’s something that we’re excited to deliver to the industry for the first time.
Notably, our CZ gate also provides compelling levels of hardware performance for near- term algorithm exploration using the Quantum Circuits error aware feature set. With a more reliable gate, one can run larger applications using more physical qubits, and explore a wider space of new, innovative approaches for algorithm development that use quantum error detection with our DRQs. This presents novel opportunities for near-term enhancements of quantum workflows, and we’re excited to be working with our partners and customers in exploring the possibilities that our rich feature set and higher fidelities can uncover.
The following points summarize our results in more detail:
- Our CZ gate has a fidelity that exceeds 99.9%, after error detection, and is thereby one of the best-performing two-qubit gates in the quantum industry. This performance is critical for efficient error correction that minimizes resource utilization.
- Detectable errors are low, occurring only about 0.5% of the time. This means we only discard one in 200 attempts to achieve this high post-selected fidelity. This is well below the limits that can be fixed with a conventional error correction code. Moreover, during the error correction cycles themselves within the enhanced version of the surface code we use, we do not discard any shots while maintaining the high levels of performance we report here.
- The gate is high fidelity and fast – 500ns. The unique combination of speed and high fidelity that Quantum Circuits’ DRQs offer is an important factor in the future practicality of quantum computing, where high clock speeds are essential to execute applications in a reasonable amount of time.
- The DRQs are high-reliability elements. The SWS protocol engineers a 5X difference in phase coherence between the two DRQs, where one reaches a dephasing benchmark upward of 5 milliseconds, even during a SWS gate. This is 10,000 times longer than the CZ duration, a ratio of coherence to speed that is very beneficial for quantum error correction.
- After DRQ error detection, we are only left with dephasing errors. Bit flips are virtually negligible in our DRQ system, reaching levels below one part in a million. Dephasing is therefore the remaining dominant error to be corrected, highlighting the remarkable power of the DRQ system design.
Our two-qubit gate results underscore how excellent fidelities can be obtained on a DRQ-based platform with integrated error detection while achieving fast gate speeds and preserving our impressive coherence times. Putting this all together means that Quantum Circuits has a platform which can beat the thresholds for fault-tolerant computing by a large margin, and scaling up to even modest-sized error correction codes will lead to even larger gains in logical qubit fidelity.
The new SWS protocol and the resulting high-performance CZ gate, with a fidelity exceeding 99.9%, presents an exciting milestone as we continue to pave the way toward scalable fault-tolerant quantum computing. Stay tuned as we release larger systems that leverage the capabilities of this gate, allowing users to execute new applications, explore new error correction techniques and break new ground with quantum systems that are performant, fast, and feature-rich.
To learn more, please download our arXiv paper.
