| CPC G06F 30/27 (2020.01) [G06F 30/398 (2020.01); G06N 3/044 (2023.01); G06N 3/08 (2013.01); G06N 3/10 (2013.01)] | 17 Claims |
|
1. A process simulation system comprising:
a random access memory (RAM) to which a process simulator is loaded, the process simulator comprising a plurality of process emulation cells operating as a time series-based recurrent neural network (RNN) and outputting profiles respectively corresponding to steps of process of manufacturing semiconductors; and
a central processing unit (CPU) configured to control the plurality of process emulation cells to train and predict the profiles based on a final target profile corresponding to a final step among the steps of process,
wherein a first process emulation cell among the plurality of process emulation cells is configured to, based on a previous output profile that is output at a previous process step in time series, a target profile of a current process step, process condition information and prior knowledge information, generate a current output profile corresponding to the current process step,
wherein the process condition information indicates one or more conditions to be applied in the current process step, and
wherein the prior knowledge information is used to filter the current output profile or limit a shape indicated by the current output profile.
|