SilTerra Shares Applications of Advanced Data Analytics in Semiconductor Manufacturing with the Malaysian Academy of Mathematical Scientists
10 Mar 2020
SilTerra Shares Applications of Advanced Data Analytics in Semiconductor Manufacturing with the Malaysian Academy of Mathematical Scientists

SilTerra was honoured to be invited by the Malaysian Academy of Mathematical Scientists to share about the applications of advanced data analytics in semiconductor fabrication. This event was held at the Institute of Mathematical Research in Universiti Putra Malaysia (UPM) on the 21st February. Through this initiative, the university students, both undergraduate and post-graduate, had the opportunity to gain exposure on the practical applications of mathematics, programming and advanced data analytics in the industry.

In his presentation, Deputy Director of Manufacturing Systems, Dr. Mohd Azizi, shared about the complexities and challenges of semiconductor manufacturing; and the utilization of advanced data analytics to solve manufacturing complexities and drive productivity.

Dr. Mohd. Azizi used real-life examples in SilTerra’s factory to explain how capacity planning is done when dealing with hundreds of tools and thousands of production lots; and the implications of production constraints such as process variability, equipment set up time, order prioritization or process dedications towards capacity planning.

The normal operation of SilTerra’s production floor requires continuous equipment monitoring. In a typical scenario, there are as many as 200 technicians and engineers overseeing wafer production and 6000 manufacturing transactions taking place per second, with over 600 active equipment that need to be maintained at an overall line yield of 99%.  Additionally, there are about 100 different chemicals and gases being supplied into the fab, with more than 1000 ancillary equipment for supplying energy, water, chemicals and data communications. At any given time, there is also an average of 60,000 wafers in the fab with their performances tracked in real-time, by the hour, by day, by lot and by shift.

All these result in over 50GB of data generated per month. With over 150TB of data in storage, the data will be used for descriptive, predictive and prescriptive reporting to address issues relating to cycle time, tool productivity, cost and yield – through analytics as well as proprietary mathematical modeling.

As one of the most advanced technology companies in Malaysia, SilTerra applies Six Sigma methodology to improve the operations of its fully integrated state-of-the-art 200mm wafer fabrication facility in Kulim Hi-Tech Park.  SilTerra is committed to nurturing talent and recognizes the importance of industry-academia collaboration for building human capital to support the E&E industry and the development of Malaysia’s high-tech ecosystem.