Error Mitigation for Universal Gates on Encoded Qubits.

The Eastin-Knill theorem states that no quantum error-correcting code can have a universal set of transversal gates. For Calderbank-Shor-Steane codes that can implement Clifford gates transversally, it suffices to provide one additional non-Clifford gate, such as the T gate, to achieve universality. Common methods to implement fault-tolerant T gates, e.g., magic state distillation, generate a significant hardware overhead that will likely prevent their practical usage in the near-term future. Recently, methods have been developed to mitigate the effect of noise in shallow quantum circuits that are not protected by error correction. Error mitigation methods require no additional hardware resources but suffer from a bad asymptotic scaling and apply only to a restricted class of quantum algorithms. In this Letter, we combine both approaches and show how to implement encoded Clifford+T circuits where Clifford gates are protected from noise by error correction while errors introduced by noisy encoded T gates are mitigated using the quasiprobability method. As a result, Clifford+T circuits with a number of T gates inversely proportional to the physical noise rate can be implemented on small error-corrected devices without magic state distillation. We argue that such circuits can be out of reach for state-of-the-art classical simulation algorithms.

Bulk and Nanocrystalline Cesium Lead-Halide Perovskites as Seen by Halide Magnetic Resonance.

Lead-halide perovskites increasingly mesmerize researchers because they exhibit a high degree of structural defects and dynamics yet nonetheless offer an outstanding (opto)electronic performance on par with the best examples of structurally stable and defect-free semiconductors. This highly unusual feature necessitates the adoption of an experimental and theoretical mindset and the reexamination of techniques that may be uniquely suited to understand these materials. Surprisingly, the suite of methods for the structural characterization of these materials does not commonly include nuclear magnetic resonance (NMR) spectroscopy. The present study showcases both the utility and versatility of halide NMR and NQR (nuclear quadrupole resonance) for probing the structure and structural dynamics of CsPbX3 (X = Cl, Br, I), in both bulk and nanocrystalline forms. The strong quadrupole couplings, which originate from the interaction between the large quadrupole moments of, e.g., the 35Cl, 79Br, and 127I nuclei, and the local electric-field gradients, are highly sensitive to subtle structural variations, both static and dynamic. The quadrupole interaction can resolve structural changes with accuracies commensurate with synchrotron X-ray diffraction and scattering. It is shown that space-averaged site-disorder is greatly enhanced in the nanocrystals compared to the bulk, while the dynamics of nuclear spin relaxation indicates enhanced structural dynamics in the nanocrystals. The findings from NMR and NQR were corroborated by ab initio molecular dynamics, which point to the role of the surface in causing the radial strain distribution and disorder. These findings showcase a great synergy between solid-state NMR or NQR and molecular dynamics simulations in shedding light on the structure of soft lead-halide semiconductors.

Mixed-Precision Deep Learning Based on Computational Memory.

Deep neural networks (DNNs) have revolutionized the field of artificial intelligence and have achieved unprecedented success in cognitive tasks such as image and speech recognition. Training of large DNNs, however, is computationally intensive and this has motivated the search for novel computing architectures targeting this application. A computational memory unit with nanoscale resistive memory devices organized in crossbar arrays could store the synaptic weights in their conductance states and perform the expensive weighted summations in place in a non-von Neumann manner. However, updating the conductance states in a reliable manner during the weight update process is a fundamental challenge that limits the training accuracy of such an implementation. Here, we propose a mixed-precision architecture that combines a computational memory unit performing the weighted summations and imprecise conductance updates with a digital processing unit that accumulates the weight updates in high precision. A combined hardware/software training experiment of a multilayer perceptron based on the proposed architecture using a phase-change memory (PCM) array achieves 97.73% test accuracy on the task of classifying handwritten digits (based on the MNIST dataset), within 0.6% of the software baseline. The architecture is further evaluated using accurate behavioral models of PCM on a wide class of networks, namely convolutional neural networks, long-short-term-memory networks, and generative-adversarial networks. Accuracies comparable to those of floating-point implementations are achieved without being constrained by the non-idealities associated with the PCM devices. A system-level study demonstrates 172 × improvement in energy efficiency of the architecture when used for training a multilayer perceptron compared with a dedicated fully digital 32-bit implementation.

Accurate deep neural network inference using computational phase-change memory.

In-memory computing using resistive memory devices is a promising non-von Neumann approach for making energy-efficient deep learning inference hardware. However, due to device variability and noise, the network needs to be trained in a specific way so that transferring the digitally trained weights to the analog resistive memory devices will not result in significant loss of accuracy. Here, we introduce a methodology to train ResNet-type convolutional neural networks that results in no appreciable accuracy loss when transferring weights to phase-change memory (PCM) devices. We also propose a compensation technique that exploits the batch normalization parameters to improve the accuracy retention over time. We achieve a classification accuracy of 93.7% on CIFAR-10 and a top-1 accuracy of 71.6% on ImageNet benchmarks after mapping the trained weights to PCM. Our hardware results on CIFAR-10 with ResNet-32 demonstrate an accuracy above 93.5% retained over a one-day period, where each of the 361,722 synaptic weights is programmed on just two PCM devices organized in a differential configuration.