APC Solution Uses AI-Driven Adaptive Intervals for Precision Probe Needle Cleans to Minimize Online Cleaning Frequency, Yield Impact
TOKYO, Japan – April 26, 2022 – Leading semiconductor test equipment supplier Advantest Corporation (TSE: 6857) today announced its industry-first ACS Adaptive Probe Cleaning (APC) solution to optimize the probe card cleaning cycle with efficacy. Part of Advantest's ACS open solution ecosystem, APC has been proven with customers throughout Asia and Europe. STMicroelectronics has installed ACS APC at two of its manufacturing sites and is seeing cleaning cycle times slashed significantly to effective improves yield, lengthen probe card life cycle and increase system available time for test.
While there are many techniques aimed at reducing test time and cost, probe cards cost presents an emerging challenge as pin count and functional complexity increase, so does the ratio of probe card expenses in both maintenance and life cycle management. Typically, chipmakers implement a fixed-cycle online cleaning method, which requires a learning curve to determine the ideal cleaning-to-yield-loss tradeoff and thus takes longer time in new device ramp and incur more loss than necessary.
ACS APC introduces Advantest's "adaptive interval" approach to probe tip cleaning. It uses artificial intelligence (AI) algorithms to assess needles' dirtiness and cleans them only when yields are being affected to drastically reduce cleaning frequency. The machine learning model takes test result from the first wafer with a pre-fixed cleaning cycle to learn the failure pattern, then it shifts to adaptive cleaning for the rest of the wafers in the same lot. The assessment takes negligible time per wafer, without impacting test throughput.
Jean-Luc Mariani, EWS Europe Operations & Technology Director and Chris Portelli-Hale, Test Technology R&D Director, both of ST Microelectronics, confirmed that they have installed several Advantest ACS APC units to aid in testing a variety of complex devices, and are seeing impressive results. They noted that their cleaning cycles have been significantly reduced, which has enabled them to cut probe card maintenance costs in half and bring about more effective probe card spare and life cycle management.
"Our new APC solution demonstrates Advantest’s Grand Design commitment to add customer value in the evolving semiconductor value chain," said Titan Chang, Executive Officer, Field Service Group, for Advantest. "Customers conducting wafer sort testing will reap the benefits of our adaptive interval technology by way of significant yield stabilization and probe card cleaning and maintenance optimization, resulting in total test cost reduction and productivity gains."
The new Advantest ACS Adaptive Probe Cleaning solution works with all existing ATE platforms and wafer probers and requires neither program modification nor use of specific parametric data. It has been proven to operate seamlessly in combination with dual probe card cleaning sheets at different cleaning frequencies.
About Advantest Corporation
Advantest (TSE: 6857) is the leading manufacturer of automatic test and measurement equipment used in the design and production of semiconductors for applications including 5G communications, the Internet of Things (IoT), autonomous vehicles, high performance computing (HPC) including artificial intelligence (AI) and machine learning, and more. Its leading-edge systems and products are integrated into the most advanced semiconductor production lines in the world. The company also conducts R&D to address emerging testing challenges and applications; develops advanced test-interface solutions for wafer sort and final test; produces scanning electron microscopes essential to photomask manufacturing; and offers system-level test solutions and other test-related accessories. Founded in Tokyo in 1954, Advantest is a global company with facilities around the world and an international commitment to sustainable practices and social responsibility. More information is available at www.advantest.com.
Note: All information supplied in this release is correct at the time of publication, but may be subject to change.