English
Related papers

Related papers: INT v.s. FP: A Comprehensive Study of Fine-Grained…

200 papers

The ever-growing cost of both training and inference for state-of-the-art neural networks has brought literature to look upon ways to cut off resources used with a minimal impact on accuracy. Using lower precision comes at the cost of…

Machine Learning · Computer Science 2021-02-03 Quentin Ducasse , Pascal Cotret , Loïc Lagadec , Robert Stewart

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…

The massive computational costs associated with large language model (LLM) pretraining have spurred great interest in reduced-precision floating-point representations to accelerate the process. As a result, the BrainFloat16 (BF16) precision…

Machine Learning · Computer Science 2025-03-26 Joonhyung Lee , Jeongin Bae , Byeongwook Kim , Se Jung Kwon , Dongsoo Lee

Using fewer bits to represent model parameters and related tensors during pre-training has become a required technique for improving GPU efficiency without sacrificing accuracy. Microscaling (MX) formats introduced in NVIDIA Blackwell…

Machine Learning · Computer Science 2025-08-20 Asit Mishra , Dusan Stosic , Simon Layton , Paulius Micikevicius

Large Language Models (LLMs) have showcased remarkable impacts across a wide spectrum of natural language processing tasks. Fine-tuning these pretrained models on downstream datasets provides further significant performance gains; however,…

Computation and Language · Computer Science 2026-03-19 Zhikai Li , Xiaoxuan Liu , Banghua Zhu , Zhen Dong , Qingyi Gu , Kurt Keutzer

Mixed-precision quantization is a promising approach for compressing large language models under tight memory budgets. However, existing mixed-precision methods typically suffer from one of two limitations: they either rely on expensive…

Machine Learning · Computer Science 2026-02-03 Xin Nie , Haicheng Zhang , Liang Dong , Beining Feng , Jinhong Weng , Guiling Sun

Deploying large language models (LLMs) on edge devices requires extremely low-bit quantization. Ultra-low precision formats such as NVFP4 offer a promising solution for reducing memory footprint and accelerating computation. However,…

Machine Learning · Computer Science 2026-03-25 Hanglin Li , Shuchang Tian , Chen Lin , Zhiyong Zhao , Kun Zhan

The increasing computational demands of foundation models have spurred research into low-precision training, with 4-bit floating-point (\texttt{FP4}) formats emerging as a frontier for maximizing hardware throughput. While numerous…

Machine Learning · Computer Science 2025-09-23 Robert Hu , Carlo Luschi , Paul Balanca

Extremely low-bit quantization is critical for efficiently deploying Large Language Models (LLMs), yet it often leads to severe performance degradation at 2 bits and even at 4 bits (e.g., MXFP4). We present SignRoundV2, a post-training…

Computation and Language · Computer Science 2026-05-19 Wenhua Cheng , Weiwei Zhang , Heng Guo , Haihao Shen , Zaner Ma

Reinforcement learning (RL) for large language models (LLMs) is increasingly bottlenecked by rollout (generation), where long output sequence lengths make attention and KV-cache memory dominate end-to-end step time. FP8 offers an attractive…

Machine Learning · Computer Science 2026-04-13 Zhaopeng Qiu , Shuang Yu , Jingqi Zhang , Shuai Zhang , Xue Huang , Jingyi Yang , Junjie Lai

Large language models have been widely adopted but require significant GPU memory for inference. We develop a procedure for Int8 matrix multiplication for feed-forward and attention projection layers in transformers, which cut the memory…

Machine Learning · Computer Science 2022-11-11 Tim Dettmers , Mike Lewis , Younes Belkada , Luke Zettlemoyer

Neural network training is a memory- and compute-intensive task. Quantization, which enables low-bitwidth formats in training, can significantly mitigate the workload. To reduce quantization error, recent methods have developed new data…

Machine Learning · Computer Science 2024-11-19 Wenjin Guo , Donglai Liu , Weiying Xie , Yunsong Li , Xuefei Ning , Zihan Meng , Shulin Zeng , Jie Lei , Zhenman Fang , Yu Wang

8-bit integer inference, as a promising direction in reducing both the latency and storage of deep neural networks, has made great progress recently. On the other hand, previous systems still rely on 32-bit floating point for certain…

Computation and Language · Computer Science 2020-09-21 Ye Lin , Yanyang Li , Tengbo Liu , Tong Xiao , Tongran Liu , Jingbo Zhu

The recent surge of interest in Deep Neural Networks (DNNs) has led to increasingly complex networks that tax computational and memory resources. Many DNNs presently use 16-bit or 32-bit floating point operations. Significant performance…

Distributed, Parallel, and Cluster Computing · Computer Science 2019-01-23 Zachariah Carmichael , Hamed F. Langroudi , Char Khazanov , Jeffrey Lillie , John L. Gustafson , Dhireesha Kudithipudi

Fast and energy-efficient low-bitwidth floating-point (FP) arithmetic is essential for Artificial Intelligence (AI) systems. Microscaling (MX) standardized formats have recently emerged as a promising alternative to baseline low-bitwidth FP…

Hardware Architecture · Computer Science 2025-05-20 Gamze İslamoğlu , Luca Bertaccini , Arpan Suravi Prasad , Francesco Conti , Angelo Garofalo , Luca Benini

Reduced-precision data formats are crucial for cost-effective serving of large language models (LLMs). While numerous reduced-precision formats have been introduced thus far, they often require intrusive modifications to the software…

Machine Learning · Computer Science 2025-10-17 Jungi Lee , Junyong Park , Soohyun Cha , Jaehoon Cho , Jaewoong Sim

Quantization-aware training (QAT) is typically performed for a single target numeric format, while practical deployments often need to choose numerical precision at inference time based on hardware support or runtime constraints. We study…

Machine Learning · Computer Science 2026-04-02 Zifei Xu , Sayeh Sharify , Hesham Mostafa

Quantizing the activation, weight, and gradient to 4-bit is promising to accelerate neural network training. However, existing 4-bit training methods require custom numerical formats which are not supported by contemporary hardware. In this…

Machine Learning · Computer Science 2023-06-26 Haocheng Xi , Changhao Li , Jianfei Chen , Jun Zhu

Post-Training Quantization (PTQ) converts pre-trained Full-Precision (FP) models into quantized versions without training. While existing methods reduce size and computational costs, they also significantly degrade performance and…

Machine Learning · Computer Science 2025-12-24 Boyang Zhang , Daning Cheng , Yunquan Zhang , Jiake Tian , Jing Li , Fangming Liu

Single-precision floating point (FP32) data format, defined by the IEEE 754 standard, is widely employed in scientific computing, signal processing, and deep learning training, where precision is critical. However, FP32 multiplication is…

Hardware Architecture · Computer Science 2025-10-09 Bindu G Gowda , Yogesh Goyal , Yash Gupta , Madhav Rao