当前位置:首页 > 报告详情

AI驱动的3DIC测试 ModelOps.pdf

上传人: 哆哆 编号:630970 2025-04-19 63页 10.33MB

1、Executive ConferenceArtificialArtificialIntelligenceIntelligenceexplore the power of AI to transform semiconductor design&manufacturingJeff David,PDF Solutions Jeff David,PDF Solutions Edward Zhou,PDF SolutionsEdward Zhou,PDF SolutionsKen Butler,AdvantestKen Butler,AdvantestAI Driven 3DIC TestAI Dri

2、ven 3DIC TestProduct and Architecture OverviewThis presentation and discussions resulting from it may include future product features or fixes,or the expected timing of future releases.This information is intended only to highlight areas of possible future development and current prioritizations.Not

3、hing in this presentation or the discussions stemming from it are a commitment to any future release,new product features or fixes,or the timing of any releases.Actual future releases may or may not include these product features or fixes,and changes to any roadmap or timeline are at the sole discre

4、tion of PDF Solutions,Inc.and may be made without any requirement for updating.For information on current prioritizations and intended future features or fixes,contact .PDF Solutions,Exensio,CV,Cimetrix,the PDF Solutions logo,and the Cimetrix logo are registered trademarks of PDF Solutions,Inc.or it

5、s subsidiaries.All other trademarks cited in this document are the property of their respective owners.Exensio visualizations Powered by TIBCO.2024 PDF Solutions,Inc.or its subsidiaries.All rights reservedCopyright PDF Solutions 2024PDF Solutions PDF Solutions ModelOpsModelOpsCreate,Manage and Contr

6、ol models across their lifecyclesCreate,Manage and Control models across their lifecycles Single platform for all data in Single platform for all data in ExensioExensio Infrastructure for semiconductor-specific dataTrain,deploy,execute,and monitor modelsTrain,deploy,execute,and monitor models Centra

word格式文档无特别注明外均可编辑修改,预览文件经过压缩,下载原文更清晰!
三个皮匠报告文库所有资源均是客户上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作商用。
本文主要介绍了PDF Solutions公司的Exensio平台,该平台是一个集成的系统,用于管理半导体行业中的数据和AI模型的整个生命周期。文章详细描述了模型的训练、部署和监控过程,并强调了模型在预测测试结果、优化生产和减少成本方面的重要性。关键数据包括:预测分拣的失败率1.00%,算法过度kill率77.80%,以及通过虚拟计量和自适应测试等方法提高的数据处理效率。此外,文章还提到了模型部署的架构,包括边缘盒子、中央训练站点和测试设备等。最后,强调了ModelOps平台在整合数据、产品化模型、与MES/ERP系统集成和实时监控模型方面的优势。
"AI如何助力半导体设计与制造?" "ModelOps平台如何优化半导体行业模型管理?" "预测性分拣技术在半导体行业中的应用有哪些优势?"
客服
商务合作
小程序
服务号
折叠