Q-CTRL Awarded £1 Million Funding in UK Quantum Catalyst Competition
February 5, 2024
A quantum leap forward for neutral-atom quantum computers has been achieved. Last October, Atom Computing announced that they had achieved over 1,000 qubits on-premises. In January, QuEra unveiled an ambitious new roadmap to put a 256-qubit device on the cloud this year. And companies like Pasqal are up and running with devices at the 100 qubit size.
These qubit counts are competitive with those achieved using other technologies, and exciting recent advances suggest the neutral-atom technology can scale even further in the near term. With two-qubit fidelities well over 99% and highly parallelizable gate operation, error-corrected algorithms are now on the agenda. Is the end of the noisy intermediate scale quantum (NISQ) era on the horizon? Quantum error correction (QEC) will take us beyond the NISQ era, so that gate errors no longer determine the accuracy of computational results. Learn about what’s required to make QEC a practical reality in our recent blog.
Q-CTRL’s infrastructure software is bringing that horizon closer. Our industry-leading product, Boulder Opal, provides “quantum EDA” capabilities for researchers pushing the limits of today’s quantum hardware and designing the platforms of tomorrow.
Using the product’s optimization functionality and tests on QuEra’s Aquila hardware, we demonstrate scalable preparation of the many-body states used for graph optimization with a 3x reduction in error relative to the standard implementation. With the same software, you can also design laser control operations that decrease two-qubit gate errors by orders of magnitude while eliminating errors from common hardware noise sources. Below we’ll show you more about what you can achieve by combining Boulder Opal with QuEra’s hardware.
Our QuEra Aquila hardware results for preparing the Z2 many-body state are shown below. The Z2 state is a commonly used “quantum wire” for linking nodes in graph optimization problems. Generating this state involves applying a shaped pulse of laser light to induce the qubit-qubit interaction. In these experiments, the laser control pulses were designed using Boulder Opal’s simulation and model-based optimization capabilities on a chain of just 7 qubits. Note that the excellent quality of state generation persists up to a 13-qubit linear chain, the largest currently available on AWS.
The optimized control yields >70% probability on the target bitstring, over 3x higher than the standard flat-top pulse.
Q-CTRL optimization tools also help you design the control pulses that implement two-qubit CZ gates using the Rydberg interaction in neutral atom quantum processors.
Using Boulder Opal, you can design pulses that improve peak gate infidelity by several orders of magnitude vs a benchmark adiabatic rapid passage (ARP) pulse.
The pulses you can design using Boulder Opal are robust to common sources of experimental error like amplitude miscalibration and detuning drifts, as shown in the image. You can also use Boulder Opal’s flexible pulse engineering features to include robustness to other error sources, such as the decay of the Rydberg state used to generate entanglement in these gates. The infidelity of these robust pulses is reduced by orders of magnitude below the benchmark over a wide range of experimental error values, as shown in the image.
When combined with easy cloud access to QuEra’s Aquila device, the optimized control techniques delivered by our quantum EDA tools can unlock much higher performance for analog operations right now. And in the future, the ability to design and deploy novel error-robust robust controls for two-qubit gates will make high-fidelity digital quantum computing a reality on hundreds of qubits with neutral atom hardware.
If you’re building the quantum platforms of the future, try Boulder Opal now or check out our application note to see how you can apply advanced simulation, control design, hardware characterization, performance verification, and AI automation capabilities to neutral atom systems.