十字线
极紫外光刻
计算机科学
德拉姆
平版印刷术
检查时间
人工智能
工程类
计算机硬件
光学
电气工程
薄脆饼
心理学
物理
发展心理学
作者
Priyank Jain,Dongxue Chen,Eric Kwon,Howard J. Huang,Jiuk Hur,Yura Chung,Sieun Jang,JuHyoung Lee,Paul Chung
摘要
EUV Lithography has grown rapidly since its first production use in 2019. The reticle quality and defect control have become of utmost importance now than ever. KLA's Teron 647e reticle inspection system and its database inspection capability have been the workhorse for mask inspection at N5 and below. To support tighter defect requirements for EUV mask inspection at advanced design nodes, KLA has pushed the envelope of DUV inspection technology and developed the next-generation Teron 647e system with advanced super-algorithms, including a hybrid database algorithm and a deep learning database algorithm. In this paper, we will discuss the evaluation results of the new reticle inspection system on critical logic and DRAM layers using a hybrid database algorithm. The next-generation reticle inspector demonstrated better sensitivity at much-improved inspectability compared with systems utilizing previous-generation algorithms. One specific area where the hybrid database algorithm showed high values was in the ability to handle Flare OPC variation between die to die, thereby significantly eliminating modeling-induced false defects.
科研通智能强力驱动
Strongly Powered by AbleSci AI