AOI Data-Driven Yield Gains & Key Semiconductor Trends
yieldWerx Semiconductor
Semiconductor test data & yield analytics, enabling better decision making & root cause analysis
Welcome to the latest edition of yieldWerx Insights!
Here's what we cover in this issue:
1) Enhancing semiconductor yield with Automated Optical Inspection data
2) Samsung's SF2 process reaches 30% yield
3) China targets U.S. tech giants amid trade tensions
4) ASML CEO sees low-cost AI models Like DeepSeek driving more demand
5) DeepSeek's AI models propel China's smart glasses industry
6) Strategic collaboration urged among Western allies in the semiconductor industry
yieldWerx Blog: Enhancing Semiconductor Yield with Automated Optical Inspection Data
Our latest blog post explores the fundamentals of optical inspection systems, the types of defects they can detect, and how yieldWerx modules help correlate wafer inspection data with parametric electrical tests—ultimately reducing yield losses.
yieldWerx provides advanced wafer mapping, inspection image management, and wafer merging solutions, enabling IDMs and OSATs to maximize the value of optical inspection data. With precise AOI feedback, you can refine initial assembly maps, ensuring alignment with your operational efficiency goals.
Teradyne and Infineon Forge Partnership to Enhance Power Testing
泰瑞达 and 英飞凌 have announced a strategic partnership aimed at advancing power testing solutions. As part of this collaboration, Teradyne will acquire a portion of Infineon's automated test equipment team based in Regensburg, Germany. This move is expected to bolster Teradyne's capabilities in power device testing and support Infineon's focus on its core competencies.
ASML CEO Sees Low-Cost AI Models Like DeepSeek Driving More Demand
ASML's CEO, Christophe Fouquet, addressed concerns about the impact of low-cost AI models, such as those developed by China's DeepSeek, on the semiconductor industry. He emphasized that while these models are more resource-efficient, they do not diminish the demand for advanced chip-making equipment.
Fouquet highlighted that the proliferation of AI applications, driven by cost-effective models, is expected to increase the overall need for high-performance chips, thereby benefiting companies like ASML that supply essential lithography machines for chip production.
SK Hynix Achieves Record Profits Amid AI-Driven Demand
SK hynix has reported unprecedented financial results, driven by a surge in demand for high-bandwidth memory (HBM) products utilized in artificial intelligence applications. For the full year, the company achieved a net profit of 19.76 trillion won (approximately $13.85 billion) and an operating profit of 23.467 trillion won ($16.41 billion). There is a 20-fold jump on a year-to-year basis for the last three months of 2024.
领英推荐
This significant growth highlights the company's strong position in the memory chip market and the increasing importance of AI technologies in driving semiconductor demand.
Samsung's SF2 Process Reaches 30% Yield, Mass Production Planned for Q4
Samsung Electronics has reportedly achieved an initial yield exceeding 30% in its SF2 (2nm) test production. The company plans to stabilize this process in the second half of 2025, with mass production of the Exynos 2600 chip slated for the fourth quarter. The Exynos 2600 is expected to power the upcoming Galaxy S26, scheduled for release in the first quarter of 2026.
Only a handful of companies have been identified as early adopters of Samsung’s 2nm process. Among them is Preferred Networks, Inc. , a Japanese AI-focused company leveraging Samsung’s SF2X process, which is optimized for AI and high-performance computing (HPC) applications. Another notable customer is Ambarella Inc , a U.S.-based AI semiconductor firm, utilizing Samsung’s SF2A process, specifically tailored for automotive applications.
China Targets U.S. Tech Giants Amid Trade Tensions
In response to ongoing trade disputes, Chinese officials are reportedly compiling a list of major U.S. technology companies, including Nvidia, Google, Apple, Broadcom, and Synopsys, for potential antitrust investigations and regulatory actions. 新思科技 already has a $35 billion acquisition of Ansys awaiting approval by the top Chinese regulator even though the transaction falls below China’s merger notification.
This strategy aims to leverage these actions in negotiations with the U.S. administration over tariffs and other trade issues.
WSJ reports that Western companies are also adopting a policy of "Anything But China".
DeepSeek's AI Models Propel China's Smart Glasses Industry to Challenge Meta
Chinese AI startup DeepSeek has introduced open-source AI models that are significantly enhancing the capabilities of domestic smart glasses manufacturers. Companies like Hangzhou-based Rokid are leveraging DeepSeek's cost-effective models to integrate advanced features such as real-time translation, navigation, and messaging into their augmented reality glasses. This development positions Chinese firms to compete more effectively with global players like Meta in the burgeoning smart glasses market.
Strategic Collaboration Urged Among Western Allies in the Semiconductor Industry
A recent analysis by the Center for European Policy Analysis (CEPA) underscores the necessity for the U.S. and European nations to strengthen collaboration in the semiconductor sector to counter China's growing influence. The report advocates for aligning tariffs and implementing targeted bans on specific semiconductor-dependent industries, such as drones and electric vehicles. It also emphasizes the importance of investing in research and development, supporting startups, and coordinating efforts with allies like South Korea, Japan, the UK, and Canada to maintain a competitive edge in semiconductor innovation and production.
We hope you enjoyed this edition of yieldWerx Insights. Let us know your views in the comments section.