analiza aiqbench
GauS introduces a novel differentiable framework for operator scheduling by utilizing Gaussian distributions to better capture the ordinal nature of time in hardware synthesis. This approach significantly reduces parameter space complexity compared to traditional categorical methods while enabling efficient gradient-based search on modern GPUs. The technique offers a flexible, first-of-its-kind formulation for pipelined scheduling problems that achieves Pareto-optimal results across various benchmarks.
GauS: Differentiable Scheduling Optimization via Gaussian Reparameterization
arXiv:2602.20427v2 Announce Type: replace Abstract: Efficient operator scheduling is a fundamental challenge in software compilation and hardware synthesis. While recent differentiable approaches have sought to replace traditional ones like exact solvers or heuristics with gradient-based search, they typically rely on categorical distributions that fail to capture the ordinal nature of time and suffer from a parameter space that scales poorly. In this paper, we propose a novel differentiable framework, GauS, that models operator scheduling as a stochastic relaxation using Gaussian distributions, which fully utilize modern parallel computing devices like GPUs. By representing schedules as continuous Gaussian variables, we successfully capture the ordinal nature of time and reduce the optimization space by orders of magnitude. Our method is highly flexible to represent various objectives and constraints, which provides the first differentiable formulation for the complex pipelined scheduling problem. We evaluate our method on a range of benchmarks, demonstrating that Gaus achieves Pareto-optimal results.
Abstract:Efficient operator scheduling is a fundamental challenge in software compilation and hardware synthesis. While recent differentiable approaches have sought to replace traditional ones like exact solvers or heuristics with gradient-based search, they typically rely on categorical distributions that fail to capture the ordinal nature of time and suffer from a parameter space that scales poorly. In this paper, we propose a novel differentiable framework, GauS, that models operator scheduling as a stochastic relaxation using Gaussian distributions, which fully utilize modern parallel computing devices like GPUs. By representing schedules as continuous Gaussian variables, we successfully capture the ordinal nature of time and reduce the optimization space by orders of magnitude. Our method is highly flexible to represent various objectives and constraints, which provides the first differentiable formulation for the complex pipelined scheduling problem. We evaluate our method on a range of benchmarks, demonstrating that Gaus achieves Pareto-optimal results.
Machine Learning (cs.LG); Hardware Architecture (cs.AR)
arXiv:2602.20427 [cs.LG]
(or arXiv:2602.20427v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2602.20427 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 242 cuvinte
Subiecte conexe
On Reparameterizing the Score Function
arXiv · 21 oct. 2026
Robust Learning under Non-Gaussian Noise with Adaptive Log-Correntropy Loss
arXiv · 4 nov. 2026
ALCL: An Adaptive Log-Correntropy Loss for Robust Learning under Non-Gaussian Noise
arXiv · 16 iun. 2026
Hardware-Native Joint Sparse-Quantization for Trillion-Scale Mixture-of-Experts
arXiv · 6 oct. 2026