Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

Interpreting ¹H NMR Signal Splitting: The (n + 1) Rule01:10

Interpreting ¹H NMR Signal Splitting: The (n + 1) Rule

1.2K
In the AX proton spin system, proton A can sense the two spin states of a coupled proton X, resulting in a doublet NMR signal with two peaks of equal (1:1) intensity. When proton A is coupled to two equivalent protons (AX2 spin system), the spin states of each X can be aligned with or against the external field, creating three possible scenarios. This results in a 1:2:1  triplet signal, where the central peak corresponds to the chemical shift of A and is twice as large or intense as the...
1.2K

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

MixTrain: accelerating DNN training via input mixing.

Frontiers in artificial intelligence·2024
Same author

Compute in-Memory with Non-Volatile Elements for Neural Networks: A Review from a Co-Design Perspective.

Advanced materials (Deerfield Beach, Fla.)·2022
Same author

Neural Network Training With Asymmetric Crosspoint Elements.

Frontiers in artificial intelligence·2022
Same author

Accelerating DNN Training Through Selective Localized Learning.

Frontiers in neuroscience·2022
Same author

Probabilistic Spike Propagation for Efficient Hardware Implementation of Spiking Neural Networks.

Frontiers in neuroscience·2021
Same author

Algorithm for Training Neural Networks on Resistive Device Arrays.

Frontiers in neuroscience·2020
Same journal

Cross-linguistic patterns of cognitive biases in large language models: a comparative study in English, Hebrew, and Russian.

Frontiers in artificial intelligence·2026
Same journal

From human-like AI to user adoption: the role of trust, attitude, and social influence in shaping behavioral intention.

Frontiers in artificial intelligence·2026
Same journal

Building large-scale English-Romanian literary translation resources with open models.

Frontiers in artificial intelligence·2026
Same journal

Editorial: GenAI in healthcare: technologies, applications and evaluation.

Frontiers in artificial intelligence·2026
Same journal

Logic, inference, understanding: cross-domain generalization for generative language models.

Frontiers in artificial intelligence·2026
Same journal

Label tree semantic losses for rich multi-class medical image segmentation.

Frontiers in artificial intelligence·2026
查看所有相关文章

相关实验视频

Updated: Jun 10, 2025

Scalable Fluidic Injector Arrays for Viral Targeting of Intact 3-D Brain Circuits
13:36

Scalable Fluidic Injector Arrays for Viral Targeting of Intact 3-D Brain Circuits

Published on: January 21, 2010

14.3K

LRMP:用于空间内存DNN加速器的混合精度层复制.

Abinand Nallathambi1, Christin David Bose1, Wilfried Haensch2

  • 1Elmore Family School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN, United States.

Frontiers in artificial intelligence
|October 21, 2024
PubMed
概括
此摘要是机器生成的。

我们介绍了LRMP,该方法结合了层复制和混合精度量化,以提高内存计算 (IMC) 加速器上的深度神经网络 (DNN) 性能. 这种方法显著降低了延迟,并增加了DNN的吞吐量,以最小的准确性损失.

关键词:
模拟加速器的模拟加速器在内存计算中的内存计算.混合整数线性编程 混合整数线性编程定量化定量化是什么强化学习是一种强化学习.

更多相关视频

Quantifying Intermembrane Distances with Serial Image Dilations
07:45

Quantifying Intermembrane Distances with Serial Image Dilations

Published on: September 28, 2018

6.4K
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

475

相关实验视频

Last Updated: Jun 10, 2025

Scalable Fluidic Injector Arrays for Viral Targeting of Intact 3-D Brain Circuits
13:36

Scalable Fluidic Injector Arrays for Viral Targeting of Intact 3-D Brain Circuits

Published on: January 21, 2010

14.3K
Quantifying Intermembrane Distances with Serial Image Dilations
07:45

Quantifying Intermembrane Distances with Serial Image Dilations

Published on: September 28, 2018

6.4K
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

475

科学领域:

  • 计算机工程 计算机工程
  • 人工智能的人工智能
  • 硬件加速器 硬件加速器

背景情况:

  • 深度神经网络 (DNN) 面临着日益增长的计算需求,推动对高效硬件解决方案的研究.
  • 使用非易失性内存 (NVM) 的内存计算 (IMC) 为通过空间并行性加速 DNN 提供了一个有前途的途径.
  • 现有的基于NVM的IMC加速器与非统一的层处理时间和区域限制作斗争,限制了DNN性能.

研究的目的:

  • 开发一种新的方法,LRMP,用于提高DNN在面积受限制的NVM基于IMC加速器上的性能.
  • 为了应对IMC架构中不均的层处理时间和高面积要求的挑战.
  • 通过共同考虑层复制和混合精度量化来优化DNN映射.

主要方法:

  • LRMP采用混合方法,结合了强化学习和混合整数线性编程.
  • 该方法智能地搜索层复制和混合精度定量化的设计空间.
  • 硬件意识模型指导搜索,密切反映了目标IMC加速器架构.

主要成果:

  • 在五个DNN基准中,LRMP表现出显著的绩效增长.
  • 实现了2.6-9.3倍的延迟减少和8-18倍的吞吐量改善.
  • 保持高精度,降解最小 (<1%).

结论:

  • LRMP有效地优化了DNN在基于NVM的IMC加速器上部署的DNN.
  • 层复制和混合精度定量化的联合应用对于性能提升至关重要.
  • 这种方法为在资源有限的硬件环境中加速DNN提供了实用解决方案.