Generative Models on Analog Hardware with Dynamics
For researchers in low-power generative modeling, this work provides a practical framework and training method to leverage analog hardware for image generation, achieving significant energy efficiency gains.
This paper introduces Analog Interaction Systems (AIS), a framework for hardware-implementable dynamical systems, and proposes mechanisms to narrow the expressivity gap with neural networks. On MNIST and Fashion-MNIST, their oscillator-based AIS achieves FID scores of 27.6 and 80.8, outperforming prior analog generative models by 3-4x with a 4-bit sparse architecture, while estimating an energy cost of 23uJ per image—a 2-orders-of-magnitude improvement over digital baselines.
Analog hardware platforms such as coupled oscillators and Analog Ising Machines naturally solve differential equations at a fraction of the energy cost of digital computation, making them attractive for low-power generative modeling, yet a fundamental mismatch exists: modern generative models assume flexible, software-defined dynamics, whereas analog hardware imposes fixed, physics-determined differential equations with limited approximation capacity. This paper introduces Analog Interaction Systems (AIS), a unified framework for hardware-implementable dynamical systems, and empirically characterizes their expressivity gap relative to neural network baselines. Two hardware-compatible mechanisms are proposed to narrow this gap - time-varying piecewise parameters and hidden physical states - and a Wasserstein GAN training procedure is developed to enable training of these models without requiring them to follow a specific trajectory. We characterize how area and power scale with connection density and precision, showing that sparse connectivity and low-bit-width quantized parameters are necessary for practical implementation, and estimate an energy cost of 23uJ per generated image for the chosen architecture, representing a 2-orders-of-magnitude improvement over digital baselines. On MNIST and Fashion-MNIST, our oscillator-based AIS achieves FID scores of 27.6 and 80.8, outperforming the best prior hardware-implementable analog generative models by 3-4x with a 4-bit sparse architecture.