Mr. Muthuramu GURUSAMY
Associate Vice President, Pre-Post Silicon, Tessolve Semiconductor
Mr. Muthuramu Gurusamy has been in semiconductor industry specializing in computer hardware/software for almost 30 years. His specialties include Test Technology Innovations, Validation and Verification Methodology, Product Research and Development. He has also been involved in University Research/Collaboration and coordinated Outsource Engineering Services and Solution. He is currently the Associate Vice President at Tessolve Semiconductor Pvt. Ltd. focusing on Pre/Post Silicon Business Development in APAC region. He holds a Bachelor of Engineering (Mechanical & Material) from Universiti Kebangsaan Malaysia and Professional Certification in Data Science Analytics from University of Science Malaysia.
Presentation Title
Accelerating Artificial Intelligence in Product Testing to Enable “Self-Healing” Test Pattern Generation
The semiconductor industry stands at a critical juncture. As Integrated Circuits (ICs) power an increasingly connected world, driving innovations in autonomous vehicles, hyperscale AI servers, and IoT ecosystems where the complexity of silicon design has fundamentally outpaced traditional verification methodologies. With transistor counts numbering in the billions and fabrication data growing into petabytes, the industry faces a "complexity crisis." Traditional test engineering, heavily reliant on manual scripting, static automation, and reactive debugging, can no longer guarantee zero-defect delivery within viable time-to-market windows. This paper, aligned with the AI for Product Testing Forum explores the transformative integration of Artificial Intelligence (AI) and Machine Learning (ML) into the semiconductor test lifecycle, proposing a necessary industry shift from static automation to dynamic, intelligent validation.
We examine how AI addresses the "data deluge" by converting vast repositories of fabrication and test data into actionable, predictive insights. Unlike conventional automation, which rigidly follows hardcoded scripts, AI-driven systems possess the capability to learn from historical silicon spins and adapt in real-time. This submission highlights three primary use cases that define this new paradigm.
This session aims to test engineers and industry professionals with the technical awareness to embrace these intelligent workflows, ensuring the next generation of silicon is not just tested, but intelligently validated for a zero-defect future.