Real-time Arbitrary Waveform Generation (AWG) is essential in various engineering and research applications, and often requires complex bespoke hardware and software. This paper introduces an AWG framework using an NVIDIA Graphics Processing Unit (GPU) and a commercially available high-speed Digital-to-Analog Converter (DAC) card, both running on a desktop personal computer (PC). The GPU accelerates the "embarrassingly" data parallel additive waveform synthesis framework for AWG, and the DAC reconstructs the generated waveform in the analog domain at high speed. The AWG framework is programmed using the developer-friendly Compute Unified Device Architecture (CUDA) runtime application programming interface from NVIDIA and is readily customizable, and scalable with additional parallel hardware. We present and characterize two different pathways for computing modulated radio-frequency (rf) waveforms: one pathway offers high-complexity simultaneous chirping of 1000 individual Nyquist-limited single-frequency tones for 35 ms at a sampling rate of 560 MB/s, and the other pathway allows simultaneous continuous chirping of 194 individual Nyquist-limited single-frequency tones at 100 MB/s, or 20 individual tones at 560 MB/s. This AWG framework is designed for fast on-the-fly rearrangement of a large stochastically-loaded optical tweezer array of single atoms or molecules into a defect-free array needed for quantum simulation and quantum computation applications.

翻译：实时任意波形生成（AWG）在工程和研究应用中至关重要，通常需要复杂的定制软硬件。本文介绍了一种基于NVIDIA图形处理器（GPU）及商用高速数模转换器（DAC）卡实现的AWG框架，二者均运行于台式个人计算机（PC）上。GPU加速了AWG中“易并行”的数据相加波形合成框架，DAC则负责将生成的波形式以高速在模拟域中重构。该AWG框架采用NVIDIA开发人员友好的计算统一设备架构（CUDA）运行时应用程序编程接口进行编程，具有易于定制及通过添加并行硬件扩展的特点。我们提出并表征了两种计算调制射频（射频）波形的不同路径：一种路径可在560 MB/s采样率下，在35 ms内对1000个独立奈奎斯特限制单频音调进行高复杂度同时啁啾；另一种路径可在100 MB/s采样率下对194个独立奈奎斯特限制单频音调进行同时连续啁啾，或在560 MB/s采样率下对20个独立音调进行同时连续啁啾。该AWG框架专为量子模拟与量子计算应用中所需的快速即时重组大型随机加载单原子或分子光镊阵列（形成无缺陷阵列）而设计。