High-Bandwidth Chiplet Interconnects for Advanced Packaging Technologies in AI/ML Applications: Challenges and Solutions

The demand for chiplet integration using 2.5D and 3D advanced packaging technologies has surged, driven by the exponential growth in computing performance required by artificial intelligence and machine learning (AI/ML). This article reviews these advanced packaging technologies and emphasizes criti...

Full description

Saved in:
Bibliographic Details
Main Authors: Shenggao Li, Mu-Shan Lin, Wei-Chih Chen, Chien-Chun Tsai
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Open Journal of the Solid-State Circuits Society
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10767590/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832586902398566400
author Shenggao Li
Mu-Shan Lin
Wei-Chih Chen
Chien-Chun Tsai
author_facet Shenggao Li
Mu-Shan Lin
Wei-Chih Chen
Chien-Chun Tsai
author_sort Shenggao Li
collection DOAJ
description The demand for chiplet integration using 2.5D and 3D advanced packaging technologies has surged, driven by the exponential growth in computing performance required by artificial intelligence and machine learning (AI/ML). This article reviews these advanced packaging technologies and emphasizes critical design considerations for high-bandwidth chiplet interconnects, which are vital for efficient integration. We address challenges related to bandwidth density, energy efficiency, electromigration, power integrity, and signal integrity. To avoid power overhead, the chiplet interconnect architecture is designed to be as simple as possible, employing a parallel data bus with forwarded clocks. However, achieving highyield manufacturing and robust performance still necessitates significant efforts in design and technology co-optimization. Despite these challenges, the semiconductor industry is poised for continued growth and innovation, driven by the possibilities unlocked by a robust chiplet ecosystem and novel 3D-IC design methodologies.
format Article
id doaj-art-fc1d69a7edd440bfaad51ac7754591f4
institution Kabale University
issn 2644-1349
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Open Journal of the Solid-State Circuits Society
spelling doaj-art-fc1d69a7edd440bfaad51ac7754591f42025-01-25T00:03:22ZengIEEEIEEE Open Journal of the Solid-State Circuits Society2644-13492024-01-01435136410.1109/OJSSCS.2024.350669410767590High-Bandwidth Chiplet Interconnects for Advanced Packaging Technologies in AI/ML Applications: Challenges and SolutionsShenggao Li0https://orcid.org/0000-0001-9296-1121Mu-Shan Lin1https://orcid.org/0000-0002-1268-1013Wei-Chih Chen2Chien-Chun Tsai3Design and Technology Platform, Taiwan Semiconductor Manufacturing Company Ltd., San Jose, CA, USADesign and Technology Platform, Taiwan Semiconductor Manufacturing Company Ltd., San Jose, CA, USADesign and Technology Platform, Taiwan Semiconductor Manufacturing Company Ltd., San Jose, CA, USADesign and Technology Platform, Taiwan Semiconductor Manufacturing Company Ltd., San Jose, CA, USAThe demand for chiplet integration using 2.5D and 3D advanced packaging technologies has surged, driven by the exponential growth in computing performance required by artificial intelligence and machine learning (AI/ML). This article reviews these advanced packaging technologies and emphasizes critical design considerations for high-bandwidth chiplet interconnects, which are vital for efficient integration. We address challenges related to bandwidth density, energy efficiency, electromigration, power integrity, and signal integrity. To avoid power overhead, the chiplet interconnect architecture is designed to be as simple as possible, employing a parallel data bus with forwarded clocks. However, achieving highyield manufacturing and robust performance still necessitates significant efforts in design and technology co-optimization. Despite these challenges, the semiconductor industry is poised for continued growth and innovation, driven by the possibilities unlocked by a robust chiplet ecosystem and novel 3D-IC design methodologies.https://ieeexplore.ieee.org/document/10767590/3Dblox3D-ICadvanced packagingartificial intelligence (AI) and computechiplet integrationenergy efficiency
spellingShingle Shenggao Li
Mu-Shan Lin
Wei-Chih Chen
Chien-Chun Tsai
High-Bandwidth Chiplet Interconnects for Advanced Packaging Technologies in AI/ML Applications: Challenges and Solutions
IEEE Open Journal of the Solid-State Circuits Society
3Dblox
3D-IC
advanced packaging
artificial intelligence (AI) and compute
chiplet integration
energy efficiency
title High-Bandwidth Chiplet Interconnects for Advanced Packaging Technologies in AI/ML Applications: Challenges and Solutions
title_full High-Bandwidth Chiplet Interconnects for Advanced Packaging Technologies in AI/ML Applications: Challenges and Solutions
title_fullStr High-Bandwidth Chiplet Interconnects for Advanced Packaging Technologies in AI/ML Applications: Challenges and Solutions
title_full_unstemmed High-Bandwidth Chiplet Interconnects for Advanced Packaging Technologies in AI/ML Applications: Challenges and Solutions
title_short High-Bandwidth Chiplet Interconnects for Advanced Packaging Technologies in AI/ML Applications: Challenges and Solutions
title_sort high bandwidth chiplet interconnects for advanced packaging technologies in ai ml applications challenges and solutions
topic 3Dblox
3D-IC
advanced packaging
artificial intelligence (AI) and compute
chiplet integration
energy efficiency
url https://ieeexplore.ieee.org/document/10767590/
work_keys_str_mv AT shenggaoli highbandwidthchipletinterconnectsforadvancedpackagingtechnologiesinaimlapplicationschallengesandsolutions
AT mushanlin highbandwidthchipletinterconnectsforadvancedpackagingtechnologiesinaimlapplicationschallengesandsolutions
AT weichihchen highbandwidthchipletinterconnectsforadvancedpackagingtechnologiesinaimlapplicationschallengesandsolutions
AT chienchuntsai highbandwidthchipletinterconnectsforadvancedpackagingtechnologiesinaimlapplicationschallengesandsolutions