Editing by Greg Gao

As the fabless chipmakers have become the fastest-growing sector in China's semiconductor industry in recent years, big data analytics is enjoying increasing popularity. It can provide meaningful and actionable insights to ramp up yields and improve manufacturing quality and productivity of fabless players, an industry expert pointed out.

Credit: Passionate in Marketing

Jia Jun, Vice President of Marketing of PDF Solutions(普迪飞) China, a provider of analytics solutions for semiconductor companies headquartered in the U.S., pointed out in an exclusive interview with JW Insights that only by maintaining a high yield rate can chip design firms make full use of production capacity and obtain higher profit, which is especially critical in the current semiconductor manufacturing capacity shortage wave. More and more IC design companies are beginning to pay attention to yield data analysis.

Industry data show that by 2020, China has 2,218 IC design companies, an increase of 24.6% from 1,780 in 2019. Statistics from the market analysis firm Qianzhan Industry Research Institute show that in 2020, China's IC design market reached $59 billion, far exceeding wafer manufacturing ($40 billion) and OSTA ($39 billion).

With the rapid development of integrated circuits technology, the feature sizes continue to shrink, the number of transistors on a microchip increases significantly, and chip performance also increases substantially. At the same time, the difficulty of the process and the cost has risen sharply. According to technology consulting company Gartner, the average IC design cost for a 16nm/14nm process chip is about $80 million, while a 7nm chip costs $271 million.

As the manufacturing costs significantly increased, yield optimization and improvement have been regarded as one of the most critical goals for semiconductor companies. The more important aspect is that, as the advanced semiconductor process continues to move forward, the yield challenge is increasingly related to the big data analysis capabilities of the foundries.

Jia noted that the lack of data analysis experience at the data analysis level is a big challenge faced by Chinese fabless chip makers. After getting the data, what data should be integrated and aligned? What analysis can help them solve the problem? Many chip design companies are often confused.

If semiconductor companies want to achieve more efficient yield management, the YMS (Yield Management System) is a popular choice. YMS is a system that integrates data management, data analysis, and professional tools. It can collect data for analysis in the process of chip design, wafer manufacturing, packaging, and testing, and help engineers quickly find the key points to improve yield, Jia added. 

Founded in 1991 and based in California, PDF Solutions provides advanced analytics solutions and differentiated data for semiconductor and electronics manufacturing to improve yield and quality. It has provided has maintained close cooperation with top-tier foundry and chip design companies, helping them enhance advanced process semiconductor yields, including services to the Chinese companies.

PDF Solutions also pays attention to the data analysis needs of semiconductor startups and small and medium-sized enterprises and launches corresponding data analysis modules for IC companies of different growth stages and scales to meet their individual needs.