Neural Architecture Search of Hybrid Models for NPU-CIM Heterogeneous AR/VR Devices

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhao, Yiwei, Li, Ziyun, Khwa, Win-San, Sun, Xiaoyu, Zhang, Sai Qian, Sarwar, Syed Shakib, Stangherlin, Kleber Hugo, Lu, Yi-Lun, Gomez, Jorge Tomas, Seo, Jae-Sun, Gibbons, Phillip B., De Salvo, Barbara, Liu, Chiao
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913540881776640
author Zhao, Yiwei
Li, Ziyun
Khwa, Win-San
Sun, Xiaoyu
Zhang, Sai Qian
Sarwar, Syed Shakib
Stangherlin, Kleber Hugo
Lu, Yi-Lun
Gomez, Jorge Tomas
Seo, Jae-Sun
Gibbons, Phillip B.
De Salvo, Barbara
Liu, Chiao
author_facet Zhao, Yiwei
Li, Ziyun
Khwa, Win-San
Sun, Xiaoyu
Zhang, Sai Qian
Sarwar, Syed Shakib
Stangherlin, Kleber Hugo
Lu, Yi-Lun
Gomez, Jorge Tomas
Seo, Jae-Sun
Gibbons, Phillip B.
De Salvo, Barbara
Liu, Chiao
contents Low-Latency and Low-Power Edge AI is essential for Virtual Reality and Augmented Reality applications. Recent advances show that hybrid models, combining convolution layers (CNN) and transformers (ViT), often achieve superior accuracy/performance tradeoff on various computer vision and machine learning (ML) tasks. However, hybrid ML models can pose system challenges for latency and energy-efficiency due to their diverse nature in dataflow and memory access patterns. In this work, we leverage the architecture heterogeneity from Neural Processing Units (NPU) and Compute-In-Memory (CIM) and perform diverse execution schemas to efficiently execute these hybrid models. We also introduce H4H-NAS, a Neural Architecture Search framework to design efficient hybrid CNN/ViT models for heterogeneous edge systems with both NPU and CIM. Our H4H-NAS approach is powered by a performance estimator built with NPU performance results measured on real silicon, and CIM performance based on industry IPs. H4H-NAS searches hybrid CNN/ViT models with fine granularity and achieves significant (up to 1.34%) top-1 accuracy improvement on ImageNet dataset. Moreover, results from our Algo/HW co-design reveal up to 56.08% overall latency and 41.72% energy improvements by introducing such heterogeneous computing over baseline solutions. The framework guides the design of hybrid network architectures and system architectures of NPU+CIM heterogeneous systems.
format Preprint
id arxiv_https___arxiv_org_abs_2410_08326
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Neural Architecture Search of Hybrid Models for NPU-CIM Heterogeneous AR/VR Devices
Zhao, Yiwei
Li, Ziyun
Khwa, Win-San
Sun, Xiaoyu
Zhang, Sai Qian
Sarwar, Syed Shakib
Stangherlin, Kleber Hugo
Lu, Yi-Lun
Gomez, Jorge Tomas
Seo, Jae-Sun
Gibbons, Phillip B.
De Salvo, Barbara
Liu, Chiao
Computer Vision and Pattern Recognition
Hardware Architecture
Machine Learning
Performance
Low-Latency and Low-Power Edge AI is essential for Virtual Reality and Augmented Reality applications. Recent advances show that hybrid models, combining convolution layers (CNN) and transformers (ViT), often achieve superior accuracy/performance tradeoff on various computer vision and machine learning (ML) tasks. However, hybrid ML models can pose system challenges for latency and energy-efficiency due to their diverse nature in dataflow and memory access patterns. In this work, we leverage the architecture heterogeneity from Neural Processing Units (NPU) and Compute-In-Memory (CIM) and perform diverse execution schemas to efficiently execute these hybrid models. We also introduce H4H-NAS, a Neural Architecture Search framework to design efficient hybrid CNN/ViT models for heterogeneous edge systems with both NPU and CIM. Our H4H-NAS approach is powered by a performance estimator built with NPU performance results measured on real silicon, and CIM performance based on industry IPs. H4H-NAS searches hybrid CNN/ViT models with fine granularity and achieves significant (up to 1.34%) top-1 accuracy improvement on ImageNet dataset. Moreover, results from our Algo/HW co-design reveal up to 56.08% overall latency and 41.72% energy improvements by introducing such heterogeneous computing over baseline solutions. The framework guides the design of hybrid network architectures and system architectures of NPU+CIM heterogeneous systems.
title Neural Architecture Search of Hybrid Models for NPU-CIM Heterogeneous AR/VR Devices
topic Computer Vision and Pattern Recognition
Hardware Architecture
Machine Learning
Performance
url https://arxiv.org/abs/2410.08326