Presentation
CANCELED - A Transfer Learning Framework for High-accurate Cross-workload Design Space Exploration of CPU
DescriptionTo perform cross-workload design space exploration of CPU, previous works implicitly transfer knowledge from source workloads and make predictions on the target one. However, they do not fully explore the transferability across workloads and their single basic prediction models limit the prediction accuracy.
In this paper, an open-source Transfer learning Ensemble Design Space Exploration framework (TrEnDSE) is proposed to perform cross-workload performance predictions. The black-box transferability is quantitatively dissected and explicitly utilized. An ensemble learning model and an uncertainty-driven iterative method are proposed to perform accurate and robust prediction. Experiments demonstrate TrEnDSE can reduce prediction error compared with the state-of-the-art.
In this paper, an open-source Transfer learning Ensemble Design Space Exploration framework (TrEnDSE) is proposed to perform cross-workload performance predictions. The black-box transferability is quantitatively dissected and explicitly utilized. An ensemble learning model and an uncertainty-driven iterative method are proposed to perform accurate and robust prediction. Experiments demonstrate TrEnDSE can reduce prediction error compared with the state-of-the-art.
Event Type
Work-in-Progress Poster
TimeWednesday, July 12th6:00pm - 7:00pm PDT
LocationLevel 2 Lobby
AI
Autonomous Systems
Cloud
Design
EDA
Embedded Systems
RISC-V
Security