Demonstrating Analog Inference on the BrainScaleS-2 Mobile System

Abstract

We present the BrainScaleS-2 mobile system as a compact analog inference engine based on the BrainScaleS-2 ASIC and demonstrate its capabilities at classifying a medical electrocardiogram dataset. The analog network core of the ASIC is utilized to perform the multiply-accumulate operations of a convolutional deep neural network. At a system power consumption of 5.6W, we measure a total energy consumption of $\mathrm {192 ~\mu \text {J} }$ for the ASIC and achieve a classification time of 276 $\mu$ s per electrocardiographic patient sample. Patients with atrial fibrillation are correctly identified with a detection rate of (93.7 ± 0.7)% at (14.0 ± 1.0)% false positives. The system is directly applicable to edge inference applications due to its small size, power envelope, and flexible I/O capabilities. It has enabled the BrainScaleS-2 ASIC to be operated reliably outside a specialized lab setting. In future applications, the system allows for a combination of conventional machine learning layers with online learning in spiking neural networks on a single neuromorphic platform.

Keywords

• Accelerator
• analog computing
• convolutional deep neural networks
• electrocardiography
• inference
• low-power