MMBench: Benchmarking End-to-End Multi-modal DNNs and Understanding Their Hardware-Software Implications

TL;DR

MMBench is a comprehensive benchmark suite designed to evaluate multi-modal DNNs, revealing their unique execution characteristics and guiding future hardware and software optimization for cloud and edge systems.

Contribution

The paper introduces MMBench, the first comprehensive benchmark suite for multi-modal DNNs, and provides an in-depth analysis of their system and hardware implications.

Findings

Multi-modal DNNs exhibit multi-stage execution and frequent synchronization.

They show high heterogeneity compared to uni-modal DNNs.

The benchmark extends to edge devices for broader applicability.

Abstract

The explosive growth of various types of big data and advances in AI technologies have catalyzed a new type of workloads called multi-modal DNNs. Multi-modal DNNs are capable of interpreting and reasoning about information from multiple modalities, making them more applicable to real-world AI scenarios. In recent research, multi-modal DNNs have outperformed the best uni-modal DNN in a wide range of distributed computing applications from traditional multimedia systems to emerging autonomous edge systems. However, despite their importance and superiority, very limited research attention has been devoted to understand the characteristics of multi-modal DNNs and their implications on current computing software/hardware platforms. Existing benchmarks either target uni-modal DNNs or only focus on the algorithm characteristics of multi-modal DNNs. There lacks representative benchmark suites that provide comprehensive system and architecture level analysis of multi-modal networks. To advance the understanding of these multi-modal DNN workloads and facilitate related research, we present MMBench, an open-source, end-to-end benchmark suite consisting of a set of real-world multi-modal DNN workloads with relevant performance metrics for evaluation. We then use MMBench to conduct an in-depth analysis on the characteristics of multi-modal DNNs. We demonstrate their unique characteristics of clear multi-stage execution, frequent synchronization and high heterogeneity, which distinguish them from conventional uni-modal DNNs. Finally, we conduct a case study and extend our benchmark to edge devices. We hope that our work can provide insights for future software/hardware design and optimization to underpin multi-modal DNNs on both cloud and edge computing platforms.