Neuro-Symbolic AI: An Emerging Class of AI Workloads and their Characterization

TL;DR

This paper analyzes the performance characteristics of neuro-symbolic AI models, highlighting their computational bottlenecks, limited parallelism, and the dominance of neural computation over symbolic processing.

Contribution

It provides the first detailed analysis of neuro-symbolic models' performance, revealing their computational bottlenecks and operational characteristics.

Findings

Symbolic models have less potential parallelism due to complex control flow.

Neural computation dominates when neural and symbolic parts are separable.

Data movement is a potential bottleneck in neuro-symbolic workloads.

Abstract

Neuro-symbolic artificial intelligence is a novel area of AI research which seeks to combine traditional rules-based AI approaches with modern deep learning techniques. Neuro-symbolic models have already demonstrated the capability to outperform state-of-the-art deep learning models in domains such as image and video reasoning. They have also been shown to obtain high accuracy with significantly less training data than traditional models. Due to the recency of the field's emergence and relative sparsity of published results, the performance characteristics of these models are not well understood. In this paper, we describe and analyze the performance characteristics of three recent neuro-symbolic models. We find that symbolic models have less potential parallelism than traditional neural models due to complex control flow and low-operational-intensity operations, such as scalar multiplication and tensor addition. However, the neural aspect of computation dominates the symbolic part in cases where they are clearly separable. We also find that data movement poses a potential bottleneck, as it does in many ML workloads.