English
Related papers

Related papers: To FP8 and Back Again: Quantifying Reduced Precisi…

200 papers

In the complex domain of large language models (LLMs), striking a balance between computational efficiency and maintaining model quality is a formidable challenge. Navigating the inherent limitations of uniform quantization, particularly…

Machine Learning · Computer Science 2023-07-24 Xiaoxia Wu , Zhewei Yao , Yuxiong He

This paper presents the first comprehensive empirical study demonstrating the efficacy of the Brain Floating Point (BFLOAT16) half-precision format for Deep Learning training across image classification, speech recognition, language…

Large language models (LLMs) have achieved impressive performance across various domains. However, the substantial hardware resources required for their training present a significant barrier to efficiency and scalability. To mitigate this…

Machine Learning · Computer Science 2025-05-05 Zhiwei Hao , Jianyuan Guo , Li Shen , Yong Luo , Han Hu , Guoxia Wang , Dianhai Yu , Yonggang Wen , Dacheng Tao

We propose LLM-FP4 for quantizing both weights and activations in large language models (LLMs) down to 4-bit floating-point values, in a post-training manner. Existing post-training quantization (PTQ) solutions are primarily integer-based…

Computation and Language · Computer Science 2024-04-30 Shih-yang Liu , Zechun Liu , Xijie Huang , Pingcheng Dong , Kwang-Ting Cheng

Recent work has shown that 8-bit floating point (FP8) can be used for efficiently training neural networks with reduced computational cost compared to training in FP32/FP16. In this work, we investigate the use of FP8 training in a…

Machine Learning · Computer Science 2025-07-31 Bokun Wang , Axel Berg , Durmus Alp Emre Acar , Chuteng Zhou

Large Language Models (LLMs) are now integral across various domains and have demonstrated impressive performance. Progress, however, rests on the premise that benchmark scores are both accurate and reproducible. We demonstrate that the…

Computation and Language · Computer Science 2025-10-28 Jiayi Yuan , Hao Li , Xinheng Ding , Wenya Xie , Yu-Jhe Li , Wentian Zhao , Kun Wan , Jing Shi , Xia Hu , Zirui Liu

Microscaling Floating-Point (MXFP) has emerged as a promising low-precision format for large language models (LLMs). Despite various post-training quantization (PTQ) algorithms being proposed, they mostly focus on integer quantization,…

Computation and Language · Computer Science 2026-01-15 Manyi Zhang , Ji-Fu Li , Zhongao Sun , Haoli Bai , Hui-Ling Zhen , Zhenhua Dong , Xianzhi Yu

Training large language models with FP8 formats offers significant efficiency gains. However, the reduced numerical precision of FP8 poses challenges for stable and accurate training. Current frameworks preserve training performance using…

Machine Learning · Computer Science 2025-12-08 Yu Zhang , Hui-Ling Zhen , Mingxuan Yuan , Bei Yu

Transformer-based models, such as BERT, have been widely applied in a wide range of natural language processing tasks. However, one inevitable side effect is that they require massive memory storage and inference cost when deployed in…

Artificial Intelligence · Computer Science 2023-12-13 Jianwei Li , Tianchi Zhang , Ian En-Hsu Yen , Dongkuan Xu

Large language models (LLMs) require immense resources for training and inference. Quantization, a technique that reduces the precision of model parameters, offers a promising solution for improving LLM efficiency and sustainability. While…

Machine Learning · Computer Science 2025-02-18 Jacob Nielsen , Peter Schneider-Kamp , Lukas Galke

The state-of-the-art hardware platforms for training Deep Neural Networks (DNNs) are moving from traditional single precision (32-bit) computations towards 16 bits of precision -- in large part due to the high energy efficiency and smaller…

Machine Learning · Computer Science 2018-12-20 Naigang Wang , Jungwook Choi , Daniel Brand , Chia-Yu Chen , Kailash Gopalakrishnan

Meeting service-level objectives (SLOs) in Large Language Models (LLMs) serving is critical, but managing the high variability in load presents a significant challenge. Recent advancements in FP8 inference, backed by native hardware…

Distributed, Parallel, and Cluster Computing · Computer Science 2026-01-16 Haeun Lee , Omin Kwon , Yeonhong Park , Jae W. Lee

Training with larger number of parameters while keeping fast iterations is an increasingly adopted strategy and trend for developing better performing Deep Neural Network (DNN) models. This necessitates increased memory footprint and…

Machine Learning · Computer Science 2020-01-17 Léopold Cambier , Anahita Bhiwandiwalla , Ting Gong , Mehran Nekuii , Oguz H Elibol , Hanlin Tang

While advancements in quantization have significantly reduced the computational costs of inference in deep learning, training still predominantly relies on complex floating-point arithmetic. Low-precision fixed-point training presents a…

Machine Learning · Computer Science 2025-10-21 Hassan Hamad , Yuou Qiu , Peter A. Beerel , Keith M. Chugg

Training large language models (LLMs) models directly in low-precision offers a way to address computational costs by improving both throughput and energy efficiency. For those purposes, NVIDIA's recent Blackwell architecture facilitates…

Fluid dynamics simulations with the lattice Boltzmann method (LBM) are very memory-intensive. Alongside reduction in memory footprint, significant performance benefits can be achieved by using FP32 (single) precision compared to FP64…

Computational Physics · Physics 2022-07-28 Moritz Lehmann , Mathias J. Krause , Giorgio Amati , Marcello Sega , Jens Harting , Stephan Gekle

Training large language models is an expensive, compute-bound process that must be repeated as models scale, algorithms improve, and new data is collected. To address this, next-generation hardware accelerators increasingly support…

Machine Learning · Computer Science 2025-06-27 Huangyuan Su , Mujin Kwun , Stephanie Gil , Sham Kakade , Nikhil Anand

Quantization is a powerful tool for accelerating large language model (LLM) inference, but the accuracy-performance trade-offs across different formats remain unclear. In this paper, we conduct the most comprehensive empirical study to…

Machine Learning · Computer Science 2026-05-27 Eldar Kurtic , Alexandre Marques , Shubhra Pandit , Mark Kurtz , Dan Alistarh

Large foundation models have become central to modern machine learning, with performance scaling predictably with model size and data. However, training and deploying such models incur substantial computational and memory costs, motivating…

FP8 formats are gaining popularity to boost the computational efficiency for training and inference of large deep learning models. Their main challenge is that a careful choice of scaling is needed to prevent degradation due to the reduced…