AI for EDA Applications
Probability Prediction
Determine the functionality of circuits by predicting the logic-1 probability of each gate under random simulation
Evaluation Metrics
Mean absolute error (MAE) between the ground truth and predicted results
Models: GCN / GAT / GraphSAGE / DeepGate2 / PolarGate
Equivalent Gate Identification
Determine the functional similarity between two gates based on their truth tables
Evaluation Metrics
Mean absolute error (MAE) of the cosine similarity between their embeddings and the similarity between the pair-wise truth tables
Models: GCN / GAT / GraphSAGE / DeepGate2 / PolarGate
Sub-Circuit Boundary Identification
Determine whether each gate is located on the boundaries of any adder in an AIG
Evaluation Metrics
Binary cross-entropy (BCE) loss of the node classification results (whether the boundary of full adder or not)
Models: Gamora / HOGA
Input Boundary Identification
Output Boundary Identification
