Applications of neural networks
View Patent ↗Software for controlling processes in a heterogeneous semiconductor manufacturing environment may include a wafer-centric database, a real-time scheduler using a neural network, and a graphical user interface displaying simulated operation of the system. These features may be employed alone or in combination to offer improved usability and computational efficiency for real time control and monitoring of a semiconductor manufacturing process. More generally, these techniques may be usefully employed in a variety of real time control systems, particularly systems requiring complex scheduling decisions or heterogeneous systems constructed of hardware from numerous independent vendors.
1. A method comprising:
connecting a plurality of nodes into a neural network, each one of the nodes represented by a programming object;
defining a condition for cloning one of the plurality of nodes into a second plurality of child nodes where the one of the plurality of nodes is configured to autonomously clone and remove child nodes and the condition includes a recognition that the one of the plurality of nodes is of insufficient complexity to process inputs; and
when the condition is met, autonomously cloning the one of the plurality of nodes into two or more child nodes.
2. The method of claim 1 wherein the condition includes a processing constraint.
3. The method of claim 1 further comprising scheduling a wafer handling operation with the neural network.
4. The method of claim 3 wherein the neural network evaluates weights for a state machine.
5. A non-transitory computer-readable storage medium storing computer executable code that, when executing on one or more computing devices, performs the steps of:
connecting a plurality of nodes into a neural network, each one of the nodes represented by a programming object;
defining a condition for cloning one of the plurality of nodes into a second plurality of child nodes where the one of the plurality of nodes is configured to autonomously clone and remove child nodes and the condition includes a recognition that the one of the plurality of nodes is of insufficient complexity to process inputs; and
when the condition is met, autonomously cloning the one of the plurality of nodes into two or more child nodes.
6. The non-transitory computer program product of claim 5 wherein the condition includes a processing constraint.
7. The non-transitory computer-readable storage medium of claim 5 wherein the code further performs the step of scheduling a wafer handling operation with the neural network.
8. The non-transitory computer-readable storage medium of claim 7 wherein the neural network evaluates weights for a state machine.