Session 1: Supply-Chain Challenges including Logistics
KEYNOTE: “Future-Proofing the Supply Chain,” Donghui Lu, Corporate VP at Micron Taiwan
The turbulence of the past two years has raised concerns about supply chain resilience. As a leading global memory and storage manufacturer, Micron is leveraging its global footprints to minimize the potential risks of the disruption of supply chain. Donghui Lu, Corporate Vice President of Frontend Operations & Head of Micron Taiwan, is going to share Micron Taiwan’s strategy and practice in shifting the supply chain, diversifying the supply base, and boosting in-country sourcing. Achieving supply chain sustainability is also another priority for Micron Taiwan to ensure a more sustainable future for all.
“Localization with Vertical Integration in Wet Chemical Supply Chain,” Chia Yoke Yin, Senior Manager, Global Front End Materials Category at Micron Technologies
High volume wafer manufacturing environment requires an effective supply chain solutions to ensure smooth operations. Under current global challenges and growing complexities, we are still posted with opportunities under such dynamics. Collaborative partnership for suppliers and end-users to ride through the uncertainty and emerge stronger. With data management, adding new perspectives to supplier manufacturing, today’s discussion would like re-evaluate supply chain complexity with vertical integration study.
Session 2: Quality, Sourcing, and Metrology Challenges
“Accelerate Value Generation with Data Collaboration for Advanced Predictive Manufacturing,” Chris Han-Adebekun, PhD, VP of Business Development at Athinia, and Vish Srinivasan, Senior Director Supplier Programs at Micron
Despite of the current slowdown, semiconductor industry’s path to 1 trillion is unwavering. Driven by the so-called digitalization of everything, our market is projected to grow by doubling the size in less than a decade. But to fully capitalize on this opportunity, there are significant hurdles to overcome. Shared issues like quality excursions, supply chain resilience, and environmental sustainability are not easily solved by individual companies. With advances in big data and machine learning, we can integrate data from multiple sources to generate new insights and uncover hidden unknowns. Successes over the last years have proven that a combination of a data sharing and analytics approach with domain expertise can accelerate the time to value generation.
In this presentation, we will share an example of an end-to-end data collaboration program with workflow from automated data onboarding, secure data sharing and finally a systematic methodology to capture and quantify value generation derived from the integrated data collaboration approach. Multiple use cases will be discussed covering examples range from electronic materials to capital equipment, with the ultimate objective to deliver accelerated time to value at scale, leading to advanced predictive manufacturing and zero-defect quality performance.
“Material Purity and Contamination Control to Enable Leading Edge Semiconductor Manufacturing,” Albert Chen, Senior Director of Technology Engagement at Entegris
Materials are the key enabler in the advancement of semiconductor technology nodes, and recently, the industry is seeing a strong acceleration in the role of materials in the semiconductor space.
To meet the ever-increasing demand for high density memory and high-speed computing, significant innovation is required in device architecture and the materials used to make these devices. The emergence of vertical scaling requires new materials with higher performance and higher levels of purity than ever before. The integration of such materials into the chip fabrication increases process complexity and makes yield ramps more challenging. With more process steps in the overall device build, speed to yield is critical.
In practice, when introducing or integrating new materials in advanced processes, unexpected complex interactions between materials and micro contamination sources in the environment or other materials are usually blamed for the delay in yield ramps. Moreover, such interactions may become more enhanced in advanced processes as the devices continue to scale down, the materials become more diversified and the architectures evolve into 3D. These new challenges have called for the need to co-optimize new materials and micro contamination solutions in the advanced processes.
This presentation will highlight some challenges in co-optimizing new materials and micro contamination solutions, and the opportunity of combining new materials with defect control to enable better devices at better yield and reliability.
“A New Paradign of Process Control Solutions,” Terry Chen, Application Team Leader at NOVA
The scaling of semiconductor devices has led to complex 3D processes and introduction of new materials that require precise dimensional and material process control below the sub-nano scale. However, traditional in-line metrologies cannot always detect all process variations. To address this challenge, new process control solutions are being developed that leverage lab-based metrology technologies, in-cell measurement, and AI.
Technologies and methodologies that have been traditionally limited to Lab usage can be developed to inline environments to measure and analyze process variations currently going undetected. This requires optimizing hardware and software for the mass production environment, as well as ensuring multiple functionalities, such as 300mm FOUP capability, fab automation, wafer & die alignment, communication with a host system high throughput, and more.
As devices shrink and abundance of material increases, it becomes increasingly challenging to correlate data on the test structures in the scribe lines with actual device performance. Therefore, fabs are moving to in-cell measurement to better correlate metrology data from cell structures with e-test or yield variation. Here, spot size and the ability to eliminate the influence of surrounding materials and structures are important.
AI solutions are being developed to extract meaningful insights from the big data generated by fabs, predict process variables, and map electrical performance. This can help overcome the limitations of individual metrology tools induced by process complexity and scaled pitches. High demand for machine learning solutions will continue to drive the metrology industry to develop new algorithms and methodology to improve efficiency, accuracy, productivity, and time to solutions for high volume manufacturing.
In this presentation, we will review some of the innovative solutions developed to address the various emerging challenges of semiconductor manufacturing, offering a new paradigm that leverages technologies previous limited to lab metrology tools, as well as in-cell measurement, and AI to deliver critical data and insights to fab engineers.