KLA Corporation Opens India R&D Centre for Process Control Innovation

KLA's Bengaluru R&D Centre Applies Indian AI Talent to Chip Yield Problems

KLA Corporation, whose inspection and metrology equipment is deployed in virtually every advanced semiconductor manufacturing facility in the world, has established a dedicated research and development centre in Bengaluru to develop the next generation of defect detection, yield management, and process control algorithms. The centre, employing over 300 engineers at opening with plans to expand to 600 by 2028, will develop machine learning and computer vision software that runs on KLA's inspection and metrology systems to identify defects, measure critical dimensions, and characterise process variation in wafer manufacturing. The R&D outputs from Bengaluru will integrate directly into KLA's commercial product development pipeline, making India a source of intellectual property for one of the semiconductor industry's most technically demanding equipment categories.

The Critical Role of Process Control in Semiconductor Manufacturing

Semiconductor manufacturing is a process of extreme precision operating at scales where natural variation in materials, equipment, and environment can cause defects invisible to the human eye that nonetheless render chips non-functional or unreliable. Process control systems, the category in which KLA is the undisputed global leader, continuously monitor the manufacturing line using optical inspection, electron beam scanning, and precision measurement tools to detect deviations from specification before they propagate across an entire wafer lot. A single missed defect class can reduce yield by tens of percent in a high-volume fab, translating to tens of millions of dollars of lost revenue per month. KLA's systems are the primary defence against this yield loss, and the effectiveness of those systems depends heavily on the quality of the defect detection algorithms and statistical models that the Bengaluru R&D centre will develop.

Machine Learning Applications and Bengaluru's Role

Machine learning has transformed defect detection in semiconductor manufacturing over the past five years, enabling KLA's systems to identify subtle and previously undetectable defect signatures that evaded classical image processing algorithms. The Bengaluru R&D centre is positioned to accelerate this ML-driven innovation by accessing India's large pool of data scientists, computer vision engineers, and machine learning researchers who have developed sophisticated skills through work in India's technology sector. KLA will train these engineers in the specific domain of semiconductor manufacturing process control, combining their machine learning expertise with the deep physics and metrology knowledge that KLA's equipment teams bring. The combination is expected to produce improvements in defect detection sensitivity, classification accuracy, and throughput that would take longer to develop at KLA's US and European R&D locations where ML talent is more constrained and expensive.

India Footprint and Dholera Service Connection

KLA's Bengaluru R&D centre joins an existing India presence that includes application engineering support and customer service operations. The expanded India organisation will also serve as the primary technical support team for the Tata Electronics Dholera fab once it reaches production, providing the kind of locally available process control expertise that is essential for effective yield ramp in a new manufacturing facility. Having KLA R&D and support engineers based in India, rather than relying on teams from the United States or Europe, reduces response time for process control issues at Dholera and builds the local knowledge base that will be valuable for any additional Indian fabs that KLA's equipment is deployed in as India's semiconductor manufacturing capacity grows.