This book is a definitive guide to improving yield and reliability in the modern semiconductor industry. Blending academic research with industrial expertise, it unveils the innovative Build-In Reliability (BIR) platform. The book consists of 20 chapters, organized into four technical parts, deep diving into operational excellence and future perspectives. It emphasizes BIR's core components--Wafer-Level Reliability (WLR) and the AI-enhanced Build-In Reliability Diagnosis System (BIRDS), a knowledge-based tool designed to enable early detection of non-conformance. Part 1 introduces the BIR framework. Subsequent parts address completeness and timeliness, which are essential for establishing robust systems to enhance yield and reliability. Part 4 discusses emerging trends, including 3D packaging technologies (e.g., CoWoS), neural network-driven reliability predictions, Prognostic Health Management (PHM) for predictive maintenance, and the strategic importance of Diversity, Equity, and Inclusion (DEI) and globalization in overcoming industry challenges. This book covers a wide range of topics, including logic devices (down to 2nm), volatile and non-volatile memory tests, laboratory management, tool design, integrated circuit (IC) qualification, supplier and consumables management, statistical process control (SPC), and a novel multi-level quality and reliability training and certification framework. It also addresses the development of exceptionally reliable automotive chips, system-level lean methodologies, and optimization strategies, with a focus on yield-reliability and performance-cost trade-offs for value engineering. Complex concepts are elucidated through 51 real-world case studies, derived from decades of industry experience. These case studies emphasize root cause analysis, timely corrective actions, and innovative methodologies. Vivid figures, flowcharts, and tables are used extensively to enhance clarity and minimize reliance on complex mathematics. Each part concludes with 20 key takeaways to reinforce core insights. An indispensable resource for practitioners, researchers, and students, this book provides actionable strategies and advanced solutions for semiconductor manufacturing. It is the only comprehensive guide available today to bridge advanced academic research with real-world industrial applications in nanoelectronics manufacturing.
本書是提升現代半導體產業產量和可靠性的權威指南。結合學術研究與產業專業知識,揭示了創新的內建可靠性(Build-In Reliability, BIR)平台。
本書共分為20章,組織成四個技術部分,深入探討運營卓越和未來展望。它強調BIR的核心組件——晶圓級可靠性(Wafer-Level Reliability, WLR)和AI增強的內建可靠性診斷系統(Build-In Reliability Diagnosis System, BIRDS),這是一個基於知識的工具,旨在實現不合格的早期檢測。
第一部分介紹了BIR框架。隨後的部分則探討了完整性和及時性,這對於建立強健系統以提升產量和可靠性至關重要。第四部分討論了新興趨勢,包括3D封裝技術(例如CoWoS)、基於神經網絡的可靠性預測、預測性維護的預測健康管理(Prognostic Health Management, PHM),以及多樣性、公平性和包容性(Diversity, Equity, and Inclusion, DEI)和全球化在克服產業挑戰中的戰略重要性。
本書涵蓋了廣泛的主題,包括邏輯元件(小至2nm)、揮發性和非揮發性記憶體測試、實驗室管理、工具設計、集成電路(Integrated Circuit, IC)認證、供應商和耗材管理、統計過程控制(Statistical Process Control, SPC),以及一個新穎的多層次質量和可靠性訓練與認證框架。它還探討了極其可靠的汽車晶片的開發、系統層級的精益方法論和優化策略,重點在於產量-可靠性和性能-成本的權衡,以實現價值工程。
複雜的概念通過51個真實案例研究進行闡明,這些案例源自數十年的行業經驗。這些案例強調了根本原因分析、及時的糾正措施和創新的方法論。生動的圖形、流程圖和表格被廣泛使用,以增強清晰度並減少對複雜數學的依賴。每個部分結尾都有20個關鍵要點,以加強核心見解。
本書是從業者、研究人員和學生不可或缺的資源,提供可行的策略和先進的解決方案,適用於半導體製造。它是目前唯一一本全面的指南,將先進的學術研究與納米電子製造中的實際工業應用相結合。
以上中文內容由 AI 翻譯