Areas Where AI Can Help in Chip and Wafer Manufacturing

Author:

June 3, 2022

Shop Now

Quality is the first thing you should consider when you buy silicon wafers. Thanks to artificial intelligence and machine learning (AI/ML), wafer and chip makers can develop efficient manufacturing processes with low defect rates.

Buy silicon wafer

Areas of Potential AI Application

Research and Design

With billions of transistors in the latest generation of chipsets, it’s only a matter of time before engineers run out of ideas to accommodate more transistors in future designs. This is where artificial intelligence can come in useful for engineers.

For instance, AI/ML can analyze existing designs and identify points that could be improved to boost next-gen models’ performance. Human analysts can look at the theoretical models and run testing and verification. With AI/ML involved in R&D, the design process becomes more efficient and less time-consuming.

Process Monitoring

Coventor’s President, Dr. David M. Fried, told Semiconductor Engineering in 2018 how each tool in the fab process maintains manufacturing data. Personnel in the production line use this data to pinpoint which part of the sequence caused the defect found during QA in an individual wafer or chip.  

However, Dr. Fried said that each tool erases information after a time to make way for new data as the manufacturing continues. Some tools upload data to the host. The host then decides which information will be kept and what will be rejected.

Combined with efficient data warehousing, however, AI/ML can track the onset of imperfections as they come in and identify possible trends in the fabrication that lead to yield losses. Machine learning could also vet existing data and decide which of these should be archived or deleted entirely.

Quality Control  

AI/ML processes can both automate and simplify the quality control phase of manufacturing. Currently, silicon wafer manufacturers rely on human operators of visual inspection equipment to spot defects in the final products.  

Unfortunately, human intervention can result in errors as well as backlogs. A person in the QA section can only check as many wafers as he can before fatigue sets in. With AI, however, visual inspection machines can be taught to recognize defective patterns until they achieve accuracy that exceeds that of human operators.

Buy silicon wafer from trusted supplier

Buy Silicon Wafer From a Trusted Supplier

Quality is what you get when you buy our silicon wafers. Contact us now, and we’ll help you get the wafers you need.

Wafer World Banner