Enregistré dans:
Détails bibliographiques
Auteur principal: Lu, Chien-Ping
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2501.02156
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866916556775096320
author Lu, Chien-Ping
author_facet Lu, Chien-Ping
contents As large-scale AI models expand, training becomes costlier and sustaining progress grows harder. Classical scaling laws (e.g., Kaplan et al. (2020), Hoffmann et al. (2022)) predict training loss from a static compute budget yet neglect time and efficiency, prompting the question: how can we balance ballooning GPU fleets with rapidly improving hardware and algorithms? We introduce the relative-loss equation, a time- and efficiency-aware framework that extends classical AI scaling laws. Our model shows that, without ongoing efficiency gains, advanced performance could demand millennia of training or unrealistically large GPU fleets. However, near-exponential progress remains achievable if the "efficiency-doubling rate" parallels Moore's Law. By formalizing this race to efficiency, we offer a quantitative roadmap for balancing front-loaded GPU investments with incremental improvements across the AI stack. Empirical trends suggest that sustained efficiency gains can push AI scaling well into the coming decade, providing a new perspective on the diminishing returns inherent in classical scaling.
format Preprint
id arxiv_https___arxiv_org_abs_2501_02156
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle The Race to Efficiency: A New Perspective on AI Scaling Laws
Lu, Chien-Ping
Machine Learning
Artificial Intelligence
Performance
As large-scale AI models expand, training becomes costlier and sustaining progress grows harder. Classical scaling laws (e.g., Kaplan et al. (2020), Hoffmann et al. (2022)) predict training loss from a static compute budget yet neglect time and efficiency, prompting the question: how can we balance ballooning GPU fleets with rapidly improving hardware and algorithms? We introduce the relative-loss equation, a time- and efficiency-aware framework that extends classical AI scaling laws. Our model shows that, without ongoing efficiency gains, advanced performance could demand millennia of training or unrealistically large GPU fleets. However, near-exponential progress remains achievable if the "efficiency-doubling rate" parallels Moore's Law. By formalizing this race to efficiency, we offer a quantitative roadmap for balancing front-loaded GPU investments with incremental improvements across the AI stack. Empirical trends suggest that sustained efficiency gains can push AI scaling well into the coming decade, providing a new perspective on the diminishing returns inherent in classical scaling.
title The Race to Efficiency: A New Perspective on AI Scaling Laws
topic Machine Learning
Artificial Intelligence
Performance
url https://arxiv.org/abs/2501.02156