Automatic heterogeneous quantization of deep neural networks for low-latency inference on the edge for particle detectors

TL;DR

This paper presents an automated, heterogeneously quantized neural network method that minimizes energy and size while maintaining high accuracy, enabling nanosecond inference on edge devices for particle detection at CERN.

Contribution

It introduces a novel automated quantization approach that optimally assigns different quantizers per layer and parameter type for efficient edge inference.

Findings

Achieved nanosecond inference latency on FPGA hardware.

Reduced resource consumption by a factor of 50.

Maintained high accuracy despite aggressive quantization.

Abstract

Although the quest for more accurate solutions is pushing deep learning research towards larger and more complex algorithms, edge devices demand efficient inference and therefore reduction in model size, latency and energy consumption. One technique to limit model size is quantization, which implies using fewer bits to represent weights and biases. Such an approach usually results in a decline in performance. Here, we introduce a method for designing optimally heterogeneously quantized versions of deep neural network models for minimum-energy, high-accuracy, nanosecond inference and fully automated deployment on chip. With a per-layer, per-parameter type automatic quantization procedure, sampling from a wide range of quantizers, model energy consumption and size are minimized while high accuracy is maintained. This is crucial for the event selection procedure in proton-proton collisions at the CERN Large Hadron Collider, where resources are strictly limited and a latency of ${\mathcal O}(1)~\mu$s is required. Nanosecond inference and a resource consumption reduced by a factor of 50 when implemented on field-programmable gate array hardware are achieved.