The examples
folder contains use cases demonstrating how to utilize CircuitOps APIs for ML applications.
These examples showcase the workflow to process Intermediate Representation (IR) tables or Labeled Property Graphs (LPGs) to build the dataset needed to train a machine learning model.
-
- Predicts arc delays of standard cells based on the input pin slew, load capacitance and cell type.
-
- It is a cross-stage timing predictor which estimates net delays before routing.
-
Timing Graph Convolutional Network (GCN)
- Leverages Graph Neural Networks (GNNs) to predict slacks at the endpoints of a design.