$STOCK UNDERGROUND$

[Robertscott]

$GTCH Based on advanced mathematical algorithm and deep learning the project will explore breakthrough approach and method for a new database engine that rapidly access to any specific area of the Integrated Circuit data. It is contemplated that the database engine will perform an initial analysis of the IC data to determine area’s priority for processing. The data then will be processed according to their density and complexity levels. The layout regions will be then processed in a fully hierarchical manner according to their internal indexing and priority weight.

In addition, unlike typical existing practice, it is expected that the proprietary mathematical algorithm will verify geometrical shapes in specific innovative method by split-and-merge segments to achieve higher speed. A neural network-based algorithm will validate the heuristic’s results to ensure on-track process. One of the main advantages of this system will be the early data process as the design moves forward. As the IC designer starts his/hers block design the system will learn the block’s data as the progresses. The database will be built incrementally over time and further shortened the final block physical verification process. The system will be working with GBT’s recent multiplanar 3D microchip methods as described in its patent.