Estudio de la eficiencia en la escalabilidad de GPUs para el entrenamiento de Inteligencia Artificial

David Cortes, Carlos Juiz, Belen Bermejo
9/3/2025
cs.LGcs.AIcs.PF

No confident code implementation found for this paper.

Abstract

Training large-scale deep learning models has become a key challenge for the scientific community and industry. While the massive use of GPUs can significantly speed up training times, this approach has a negative impact on efficiency. In this article, we present a detailed analysis of the times reported by MLPerf Training v4.1 on four workloads: BERT, Llama2 LoRA, RetinaNet, and Stable Diffusion, showing that there are configurations that optimise the relationship between performance, GPU usage, and efficiency. The results point to a break-even point that allows training times to be reduced while maximising efficiency.

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

Breaking MLPerf Training: A Case Study on Optimizing BERT

84% match2024

Towards Universal Performance Modeling for Machine Learning Training on Multi-GPU Platforms

84% match2024

Forecasting GPU Performance for Deep Learning Training and Inference

83% match2024

Improving Automatic Parallel Training via Balanced Memory Workload Optimization

82% match1 repo1852023

Performance of Small Language Model Pretraining on FABRIC: An Empirical Study

82% match2026

Characterizing the Efficiency of Distributed Training: A Power, Performance, and Thermal Perspective

82% match2025

Cite this paper

@article{cortes2025estudio,
  title  = {Estudio de la eficiencia en la escalabilidad de GPUs para el entrenamiento de Inteligencia Artificial},
  author = {David Cortes and Carlos Juiz and Belen Bermejo},
  year   = {2025},
  eprint = {2509.03263},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2509.03263v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2509.03263.svg)](https://codeofpaper.com/paper/2509.03263)

Discussion