PI-DLinear integrates derived thermal ODEs into DLinear to forecast AI data center power more accurately than SOTA models while respecting physical constraints under throttling and transients.
McClenny and Ulisses M
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
method 1
citation-polarity summary
years
2026 4verdicts
UNVERDICTED 4roles
method 1polarities
use method 1representative citing papers
SPBM extends classical penalty-barrier methods to stochastic non-convex non-smooth settings via exponential dual averaging and Moreau envelopes, matching baselines with linear overhead up to 10,000 constraints.
A PINN-based periodic CFD solver is shown to reach nearly the same accuracy as traditional transient-to-periodic methods but with substantially lower computational time for 2D heat diffusion and fluid flow cases.
DDS-PINN uses localized neural networks plus a unified global loss to model multiscale fluid flows with long-range dependencies, achieving CFD-comparable accuracy on laminar backward-facing step flow with zero data and O(10^-4) error on turbulent flow with only 500 supervision points.
citing papers explorer
-
A Physics-Aware Framework for Short-Term GPU Power Forecasting of AI Data Centers
PI-DLinear integrates derived thermal ODEs into DLinear to forecast AI data center power more accurately than SOTA models while respecting physical constraints under throttling and transients.
-
Stochastic Penalty-Barrier Methods for Constrained Machine Learning
SPBM extends classical penalty-barrier methods to stochastic non-convex non-smooth settings via exponential dual averaging and Moreau envelopes, matching baselines with linear overhead up to 10,000 constraints.
-
Physics Informed Neural Network-based Computational Method for Accelerating Time-Periodic Unsteady CFD Simulations
A PINN-based periodic CFD solver is shown to reach nearly the same accuracy as traditional transient-to-periodic methods but with substantially lower computational time for 2D heat diffusion and fluid flow cases.
-
Multiscale Physics-Informed Neural Network for Complex Fluid Flows with Long-Range Dependencies
DDS-PINN uses localized neural networks plus a unified global loss to model multiscale fluid flows with long-range dependencies, achieving CFD-comparable accuracy on laminar backward-facing step flow with zero data and O(10^-4) error on turbulent flow with only 500 supervision points.