PiPSim: A Behavior-Level Modeling Tool for CNN Processing-in-Pixel Accelerators

Convolutional neural networks (CNNs) have been gaining popularity in recent years, and researchers have designed specialized architectures to speed up the inference process. However, despite the promising potential of processing near-/in-sensor architectures actively explored in the visual Internet...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on computer-aided design of integrated circuits and systems 2024-01, Vol.43 (1), p.1-1
Main Authors: Roohi, Arman, Tabrizchi, Sepehr, Morsali, Mehrdad, Pan, David Z., Angizi, Shaahin
Format: Article
Language:English
Subjects:
Accelerators
Accuracy
Algorithms
Artificial neural networks
Business metrics
Circuits
Computational modeling
Computer architecture
Computer simulation
Design optimization
energy efficiency
Integrated circuit modeling
Internet of Things
Memory management
neural network
Neural networks
Nonvolatile memory
numerical simulation
Performance evaluation
Performance measurement
Pixels
processing-in-sensor
Random access memory
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Convolutional neural networks (CNNs) have been gaining popularity in recent years, and researchers have designed specialized architectures to speed up the inference process. However, despite the promising potential of processing near-/in-sensor architectures actively explored in the visual Internet of Things, there is still a need to develop a behavior-level simulator to model performance and facilitate early design exploration. This paper proposes a stand-alone simulation platform for processing-in-pixel (PiP) systems, namely PiPSim. It offers a flexible interface and a wide range of design options for customizing the efficiency and accuracy of PiP-based accelerators using a hierarchical structure. Its organization spans from the device level, e.g., memory technology, upwards to the circuit level, e.g., compute-add on architecture, and then to the algorithm level, e.g., DNN workloads. PiPSim realizes instruction-accurate evaluation of circuit-level performance metrics as well as learning accuracy at run-time. Compared to SPICE simulation, PiPSim achieves over 25,000× speed-up with less than a 2.5% error rate on average. Furthermore, PiPSim can optimize the design and estimate the tradeoff relationships among different performance metrics.
ISSN:0278-0070
1937-4151
DOI:10.1109/TCAD.2023.3305574