News: New Applied Materials tools use AI to catch mistakes on chips.
SAN FRANCISCO (Reuters) – A new semiconductor manufacturing technology from Applied Materials uses artificial intelligence (AI) to more effectively detect defects in chips, the US company said on Tuesday.
Modern chip manufacturing takes place in factories, each costing over $ 18 billion to build and requiring hundreds of separate steps. For companies like Intel Corp., Taiwan Semiconductor Manufacturing Co., and Samsung Electronics Co. Ltd. It is crucial that chips roll off the assembly line without errors in functions that are only a few nanometers wide, without errors occurring.
The new tools applied aim to inspect these chips at different points in time during the manufacturing process. A new optical scanner – essentially an extremely advanced camera called Applied Enlight – quickly scans a silicon wafer for problem areas for 15 minutes, and then an electron microscope zooms in to take a closer look.
The problem that AI seeks to solve is that electron microscopes are accurate but slow. An initial optical scan could find a million possible problem areas on a silicon wafer, and it would take an electron microscope to examine each of those areas for days – and much of that time would be wasted given that there are only a fraction of what chip industry veterans consider Designating “killer” defects would cause the chip to malfunction.
The new artificial intelligence technology Applied ExtractAI calls only needs to examine about 1,000 of these potential problem areas with an electron microscope to predict where the biggest problems will be. Keith Wells, group vice president and general manager of imaging and process control at Applied, said the AI-driven review only takes about an hour.
“It is economical for the customer to do this on every wafer,” Wells said in an interview. “We tell you with great confidence that these are the truly fatal shortcomings.”
Applied has been testing the system with customers since last year and has generated sales of more than $ 400 million to date.
Reporting by Stephen Nellis; Adaptation by Rosalba O’Brien
Original Source © Reuters