English
Related papers

Related papers: Low-Precision Floating-Point Schemes for Neural Ne…

200 papers

In recent years, half precision floating-point arithmetic has gained wide support in hardware and software stack thanks to the advance of artificial intelligence and machine learning applications. Operating at half precision can…

Numerical Analysis · Mathematics 2024-09-19 Longfei Gao , Kevin Harms

Artificial Intelligence algorithms have been steadily increasing in popularity and usage. Deep Learning, allows neural networks to be trained using huge datasets and also removes the need for human extracted features, as it automates the…

Neural and Evolutionary Computing · Computer Science 2020-05-11 Vasco Lopes , Paulo Fazendeiro

Deep learning has grown rapidly thanks to its state-of-the-art performance across a wide range of real-world applications. While neural networks have been trained using IEEE-754 binary32 arithmetic, the rapid growth of computational demands…

Artificial Intelligence · Computer Science 2021-04-08 Valentina Popescu , Abhinav Venigalla , Di Wu , Robert Schreiber

A new deterministic floating-point arithmetic called precision arithmetic is developed to track precision for arithmetic calculations. It uses a novel rounding scheme to avoid excessive rounding error propagation of conventional…

Discrete Mathematics · Computer Science 2025-10-20 Chengpu Wang

Large-scale deep neural networks (DNN) have been successfully used in a number of tasks from image recognition to natural language processing. They are trained using large training sets on large models, making them computationally and…

Machine Learning · Computer Science 2017-03-28 Sek Chai , Aswin Raghavan , David Zhang , Mohamed Amer , Tim Shields

Convolutional neural networks (CNNs) are typically trained using 16- or 32-bit floating-point (FP) and researchers show that low-precision floating-point (FP) can be highly effective for inference. Low-precision FP can be implemented in…

Hardware Architecture · Computer Science 2021-03-02 James Garland , David Gregg

The study of the expressive power of neural networks has investigated the fundamental limits of neural networks. Most existing results assume real-valued inputs and parameters as well as exact operations during the evaluation of neural…

Machine Learning · Computer Science 2024-07-17 Yeachan Park , Geonho Hwang , Wonyeol Lee , Sejun Park

Scientific computing applications, such as computational fluid dynamics and climate modeling, typically rely on 64-bit double-precision floating-point operations, which are extremely costly in terms of computation, memory, and energy. While…

Hardware Architecture · Computer Science 2024-09-24 Cong "Callie" Hao

Information loss in numerical physics simulations can arise from various sources when solving discretized partial differential equations. In particular, errors related to numerical precision ("sub-precision errors") can accumulate in the…

Fluid Dynamics · Physics 2022-09-27 Akash Haridas , Nagabhushana Rao Vadlamani , Yuki Minamoto

Low-precision arithmetic trains deep learning models using less energy, less memory and less time. However, we pay a price for the savings: lower precision may yield larger round-off error and hence larger prediction error. As applications…

Machine Learning · Computer Science 2022-03-18 Chengrun Yang , Ziyang Wu , Jerry Chee , Christopher De Sa , Madeleine Udell

Very deep convolutional networks with hundreds of layers have led to significant reductions in error on competitive benchmarks. Although the unmatched expressiveness of the many layers can be highly desirable at test time, training very…

Machine Learning · Computer Science 2016-08-01 Gao Huang , Yu Sun , Zhuang Liu , Daniel Sedra , Kilian Weinberger

The posit number system is arguably the most promising and discussed topic in Arithmetic nowadays. The recent breakthroughs claimed by the format proposed by John L. Gustafson have put posits in the spotlight. In this work, we first…

Computer Vision and Pattern Recognition · Computer Science 2019-07-10 Raúl Murillo Montero , Alberto A. Del Barrio , Guillermo Botella

Low precision training can significantly reduce the computational overhead of training deep neural networks (DNNs). Though many such techniques exist, cyclic precision training (CPT), which dynamically adjusts precision throughout training…

Machine Learning · Computer Science 2024-03-05 Cameron R. Wolfe , Anastasios Kyrillidis

Motivated by the increasing interest in the posit numeric format, in this paper we evaluate the accuracy and efficiency of posit arithmetic in contrast to the traditional IEEE 754 32-bit floating-point (FP32) arithmetic. We first design and…

Hardware Architecture · Computer Science 2021-09-20 Stefan Dan Ciocirlan , Dumitrel Loghin , Lavanya Ramapantulu , Nicolae Tapus , Yong Meng Teo

Conventional hardware-friendly quantization methods, such as fixed-point or integer, tend to perform poorly at very low word sizes as their shrinking dynamic ranges cannot adequately capture the wide data distributions commonly seen in…

Machine Learning · Computer Science 2020-02-12 Thierry Tambe , En-Yu Yang , Zishen Wan , Yuntian Deng , Vijay Janapa Reddi , Alexander Rush , David Brooks , Gu-Yeon Wei

In low-latency or mobile applications, lower computation complexity, lower memory footprint and better energy efficiency are desired. Many prior works address this need by removing redundant parameters. Parameter quantization replaces…

Machine Learning · Computer Science 2021-11-16 Cheng-Chou Lan

Currently, deep neural networks are deployed on low-power portable devices by first training a full-precision model using powerful hardware, and then deriving a corresponding low-precision model for efficient inference on such systems.…

Machine Learning · Computer Science 2017-11-15 Hao Li , Soham De , Zheng Xu , Christoph Studer , Hanan Samet , Tom Goldstein

Quantization-aware training (QAT) is a common paradigm for network quantization, in which the training phase incorporates the simulation of the low-precision computation to optimize the quantization parameters in alignment with the task…

Machine Learning · Computer Science 2024-12-23 Chengting Yu , Shu Yang , Fengzhao Zhang , Hanzhi Ma , Aili Wang , Er-Ping Li

We propose DoReFa-Net, a method to train convolutional neural networks that have low bitwidth weights and activations using low bitwidth parameter gradients. In particular, during backward pass, parameter gradients are stochastically…

Neural and Evolutionary Computing · Computer Science 2018-02-05 Shuchang Zhou , Yuxin Wu , Zekun Ni , Xinyu Zhou , He Wen , Yuheng Zou

Modern deep neural network (DNN) models generally require a huge amount of weight and activation values to achieve good inference outcomes. Those data inevitably demand a massive off-chip memory capacity/bandwidth, and the situation gets…

Machine Learning · Computer Science 2021-04-27 Cheng-Wei Huang , Tim-Wei Chen , Juinn-Dar Huang
‹ Prev 1 4 5 6 7 8 10 Next ›