Pretraining on 16nm data and fine-tuning on 5nm data with silicon odometer features reduces Vmin prediction RMSE on 5nm chips to 3.89 mV versus 4.59 to 7.09 mV for baselines.
Lattice regression,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Transfer Learning for Minimum Operating Voltage Prediction in Advanced Technology Nodes: Leveraging Legacy Data and Silicon Odometer Sensing
Pretraining on 16nm data and fine-tuning on 5nm data with silicon odometer features reduces Vmin prediction RMSE on 5nm chips to 3.89 mV versus 4.59 to 7.09 mV for baselines.