Call for change

As semiconductor companies struggle to develop more complex chip customization and growing demand a world-leading chipmaker in the memory space was looking to secure a competitive advantage at every step of the manufacturing process. This advancement in the manufacturing process would evolve their next generation of innovative chip technology. With the goal of applying artificial intelligence and machine learning it would help deliver value and proof of concept for future product extensions.

When tech meets human ingenuity

Part of the semiconductor manufacturing process is called Chemical Mechanical Planarization (CMP), this is a core step but does not operate in real time or detect root causes.

By deploying the Semiconductor Manufacturing Analytics in the CMP process improving the speed and visibility in an advanced new chip and detect scratches. Combining trained AI models and the clients own custom edge hardware resulted in a predictive model that boosted real-time decision making and increase accuracy.

Semiconductor manufacturers can get to real time in quality control and production decisions, the better they can master their emerging market opportunities.

A valuable difference

Within three months, by harnessing AI and ML to detect and understand root causality in the CMP process, Semiconductor Manufacturing Analytics delivered significant business value for the client. The analytics concepts at the fab allowed for a way to reinvent a platform across the enterprise and leverage data as a competitive differentiator.

80%

Accuracy improvement from baseline 55% for the ensemble model.

30%

Improvement of deep learning with dense AE inference.

46%

Increased defect detection accuracy, enabling cost savings of $10 million.

Meet the team

Subscription Center
Stay in the know with our newsletter Stay in the know with our newsletter