A physics-informed residual neural network trained on per-unit power labels cuts power estimation error by 74 to 85 percent over the unsupervised ABPI baseline on Jetson hardware and a simulated heterogeneous SoC.
An evaluation framework for dynamic thermal management strategies in 3d multiprocessor system-on- chip co-design,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.PF 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
CPINN-ABPI: Physics-Informed Neural Networks for Accurate Power Estimation in MPSoCs
A physics-informed residual neural network trained on per-unit power labels cuts power estimation error by 74 to 85 percent over the unsupervised ABPI baseline on Jetson hardware and a simulated heterogeneous SoC.