可制造性设计
互连
软件部署
可扩展性
计算机科学
方案(数学)
分布式计算
带宽(计算)
计算机体系结构
直接结合
引线键合
钥匙(锁)
材料科学
电子工程
工作(物理)
集成电路封装
系统工程
热的
对称多处理机系统
硅
压缩(物理)
多边形网格
计算科学
系统集成
嵌入式系统
物理设计
过程集成
三维集成电路
机械工程
碳化硅
上市时间
集成电路
作者
Krutikesh Sahoo,Golam Sabbir,Vineeth Harish,Jui-Han Liu,Subramanian S. Iyer
出处
期刊:IEEE Transactions on Components, Packaging and Manufacturing Technology
[Institute of Electrical and Electronics Engineers]
日期:2026-02-17
卷期号:16 (4): 736-748
标识
DOI:10.1109/tcpmt.2026.3665683
摘要
Wafer-scale integration is a promising pathway to meet the increasing processor-to-memory bandwidth and memory capacity demands of next-generation high-performance computing (HPC) and artificial intelligence (AI) applications. While several architectural concepts for wafer-scale computing have been proposed, comparatively fewer efforts have focused on the practical challenges of assembly and implementation at scale. In this work, we evaluate the technological readiness of Copper-Copper (Cu–Cu) direct thermal compression bonding for enabling wafer-scale integration on a silicon interconnect fabric (Si-IF) platform. We present a high-throughput Cu–Cu bonding scheme tailored for heterogeneous integration, emphasizing key design, process, and tooling parameters that critically impact assembly yield. Strategies for further enhancing yield and manufacturability are also discussed. Our work demonstrates that Cu–Cu direct bonding on the Si-IF platform, provides a viable and scalable solution for realizing wafer-scale systems, potentially accelerating the deployment of heterogeneous wafer-scale computing in future advanced packaging ecosystems.
科研通智能强力驱动
Strongly Powered by AbleSci AI