Capturing data is costly using traditional test process software: Unwieldy volumes of data are collected, which are difficult to sift and assimilate. Other barriers to using collected data have been complexity, accessibility and integrity. But OptimalTest has combined modern database technology with adaptive testing techniques to forge a powerful new way to cull and highlight meaningful data for immediate, high-impact decision-making. Modern database management techniques make accessing the data easier for multiple diverse users, and data-mining/data analysis techniques can now quickly turn piles of raw data into valuable information in real time and nearly real time. OptimalTest’s advanced integrated infrastructure and centralized database management gives you the ability to capture data anywhere in the design-fab-test process, as well as anywhere globally, and to overlay and mine it for its value content to drive adaptive testing.

Leveraging Device Data Across the IC Lifecycle and the Enterprise. For Lower Cost of Test, Optimized Yield Learning and Augmented Reliability

Our solutions provide innovative adaptive testing techniques focused on data: capturing, centralizing, storing, moving, mining, stacking, analyzing, learning from it and – most important – making timely decisions based on it. To make you more competitive, we give you actionable data – valuable information for decision making – by enabling seamless data exchange across an enterprise, real time and off line, wafer sort and final test.

At the same time, OptimalTest’s unique implementation of advanced adaptive testing optimizes test flow by augmenting or reducing tests based on reference die results and derived criteria. This is possible because it captures and centralizes all information about the health of the device under test, not just results from the tester alone. Our approach records everything known about a device: neighborhood, wafer, lot, batch. Leveraging device data – data feed forward (DFF) and data feed backward (DFB) from fabrication and wafer sort to final test (see below) -- occurs seamlessly across the enterprise.

This multi-dimensional analysis allows the test program to be adapted and optimized based on everything that is known about the device under test, providing valuable leverage for yield and reliability learning. Data Feed Backward (DFB) can then start with automated decisions as to which chips provide the most effective data for analysis for tuning your lithography tools, optimizing your defect sampling rate and avoiding excessive data logging. This also allows for increases in capacity of installed test equipment Historically, it was difficult to justify changes in test flow due to real time capabilities and to the high demand for volume data to show impacts on outgoing quality at DPM levels. However, having the capability to simulate test flow changes using historical and real-time data can provide the business justifications needed for aggressive cost reductions with a clear understanding of their impact and an ability to record changes to align them with your yield analysis tools.