SEMI Global Standards Summit

Factories are increasing their use of digital twins, predictive maintenance, and AI/ML to improve efficiency and throughput across the entire manufacturing environment. To take full advantage of these approaches, factories must reduce cybersecurity risks and secure the transfer of 'smart' data across the entire supply chain while protecting IP. There is a need for standards to address these risk areas, as well as help diverse advanced analytics systems collaborate and interoperate to assist personnel in increasing factory productivity. Examples include adjusting processing recipes and scheduling flow-oriented processes more efficiently.

Facilitator: Albert Fuchigami (PEER Group)

The semiconductor industry is rapidly advancing towards autonomous fabrication facilities, with equipment control as the linchpin of this transformation. This presentation explores the cutting-edge developments in equipment control for autonomous fabs, highlighting the transformative role of Generative AI (Gen AI) and the critical importance of data standardization in this evolution. 

At the core of autonomous fab operations is the need for sophisticated, AI-driven equipment control systems. We'll examine how traditional control mechanisms are being revolutionized by the integration of advanced AI algorithms, particularly Gen AI. The discussion will showcase how these technologies are enabling more precise, adaptive, and predictive control of semiconductor manufacturing equipment, leading to unprecedented levels of efficiency and yield optimization. 

A key focus will be on how Gen AI is reshaping equipment control paradigms in autonomous fabs. We'll explore real-world applications where Gen AI is being used to generate optimal process recipes, enhance real-time decision-making, and even predict and mitigate potential equipment failures before they occur. The presentation will delve into how Gen AI models are trained on vast amounts of historical and real-time data to continually improve equipment performance and adapt to changing manufacturing conditions. 

While equipment control is at the forefront, we'll also address the crucial role of data standardization in enabling these advancements. We'll discuss how standardized data formats and protocols are essential for seamless integration of diverse equipment and AI systems in autonomous fabs. This includes exploring ongoing initiatives in data standardization for semiconductor manufacturing and proposing new avenues for industry-wide collaboration to develop comprehensive data standards that support both current AI applications and future Gen AI innovations in equipment control. 

The presentation will address the challenges and opportunities in implementing Gen AI for equipment control, including the need for explainable AI decisions, ensuring robustness and reliability, and maintaining human oversight in critical processes. We'll explore strategies for balancing the autonomy of AI-controlled equipment with the need for human expertise and intervention when necessary. 

Looking to the future, we'll discuss the potential of Gen AI to enable self-optimizing and self-configuring equipment in fully autonomous fabs. This includes exploring concepts like digital twins for equipment simulation and the potential for Gen AI to assist in equipment design and configuration for optimal performance in specific manufacturing scenarios.

Packaging technologies have been progressing since the beginning stages of semiconductor device development more than 70 years ago. Where packaging occurs in the semiconductor process has been changing, some of the packaging processes are now done as an extension of the front-end. The architecture and the packaging materials will be more important from now on. In this session we will discuss what kind of standardization our industry needs for the proper and healthy improvement of semiconductor manufacturing and packaging.

Facilitator: Kazunori Kato (AiT)

The presentation will review the Semiconductor Assembly Test Automation and Standardization (SATAS) Research Association goals for standards that will support and enable a fully automated assembly and test manufacturing. Existing standards, current standards development efforts as well as the SATAS standards roadmap and future vision will be reviewed.

The semiconductor industry is heading into an era of NetZero, in which quantification of environmental performance can have meaningful financial impact. The methods of measuring and accounting the environmental impact such as carbon emissions and the presence of substances of concern in manufacturing and products are not uniformly consistent across the industry. In order for the semiconductor industry to better navigate in this arena a consistent set of standards will be foundational.

Facilitator: Benjamin Gross (Applied Materials)

While the Summit highlights the standards needed for future factories, packaging architecture & materials and environmental sustainability, these elements are often interdependent bringing additional layers of complexity. The panel session will invite Summit presenters to explore these cross-cutting topics and discuss how standardization can help address these complex challenges.