More Web Proxy on the site http://driver.im/

research-article

StreamBrain: An HPC Framework for Brain-like Neural Networks on CPUs, GPUs and FPGAs

Authors:

Steven W. D. Chien,

Naresh Balaji Ravichandran,

Anders Lansner,

Stefano MarkidisAuthors Info & Claims

HEART '21: Proceedings of the 11th International Symposium on Highly Efficient Accelerators and Reconfigurable Technologies

Article No.: 8, Pages 1 - 6

https://doi.org/10.1145/3468044.3468052

Published: 21 June 2021 Publication History

Abstract

The modern deep learning method based on backpropagation has surged in popularity and has been used in multiple domains and application areas. At the same time, there are other - less-known - machine learning algorithms with a mature and solid theoretical foundation whose performance remains unexplored. One such example is the brain-like Bayesian Confidence Propagation Neural Network (BCPNN). In this paper, we introduce StreamBrain--a framework that allows neural networks based on BCPNN to be practically deployed in High-Performance Computing systems. StreamBrain is a domain-specific language (DSL), similar in concept to existing machine learning (ML) frameworks, and supports backends for CPUs, GPUs, and even FPGAs. We empirically demonstrate that StreamBrain can train the well-known ML benchmark dataset MNIST within seconds, and we are the first to demonstrate BCPNN on STL-10 size networks. We also show how StreamBrain can be used to train with custom floating-point formats and illustrate the impact of using different bfloat variations on BCPNN using FPGAs.

References

[1]

Bekolay et al. 2014. Nengo: a Python Tool for Building Large-scale Functional Brain Models. Frontiers in Neuroinformatics (2014).

[2]

Ben-Nun et al. 2019. Demystifying Parallel and Distributed Deep Learning. ACM CSUR'19 (2019).

Digital Library

[3]

Chandra et al. 2001. Parallel programming in OpenMP. Morgan Kaufmann.

Digital Library

[4]

Coates et al. 2011. An Analysis of Single-layer Networks in Unsupervised Feature Learning. In AISTATS'11.

[5]

Czajkowski et al. 2012. From OpenCL to High-Performance Hardware on FPGAs. In in FPL'12.

[6]

Davison et al. 2009. PyNN: a Common Interface for Neuronal Network Simulators. Frontiers in Neuroinformatics (2009).

[7]

De Dinechin et al. 2011. Designing Custom Arithmetic Data Paths with FloPoCo. Design & Test of Computers (2011).

Digital Library

[8]

Zohouri et al. 2018. High-performance High-order Stencil computation on FPGAs using OpenCL. In IPDPSW'18.

[9]

Fiebig et al. 2017. A Spiking Working Memory Model based on Hebbian Short-term Potentiation. Journal of Neuroscience (2017).

[10]

Fiebig et al. 2020. An Indexing Theory for Working Memory based on Fast Hebbian Plasticity. ENeuro (2020).

[11]

Gewaltig et al. 2007. Nest (Neural Simulation Tool). Scholarpedia (2007).

[12]

Gulli et al. 2017. Deep Learning with Keras.

Digital Library

[13]

Gustafson et al. 2017. Beating Floating Point at its own Game: Posit Arithmetic. Supercomputing Frontiers and Innovations (2017).

Digital Library

[14]

Hines et al. 1997. The NEURON simulation environment. Neural Computation (1997).

Digital Library

[15]

Johansson et al. 2007. Towards Cortex Sized Artificial Neural Systems. Neural Networks (2007).

Digital Library

[16]

Knight et al. 2016. Large-scale Simulations of Plastic Neural Networks on Neuromorphic Hardware. Frontiers in Neuroanatomy (2016).

[17]

Krizhevsky et al. 2012. ImageNET Classification with Deep Convolutional Neural Networks. In in NIPS'12.

Digital Library

[18]

LeCun et al. 2015. Deep learning. Nature (2015).

[19]

LeCun, Yann. 1998. The MNIST Database of Handwritten Digits. http://yann.lecun.com/exdb/mnist/ (1998).

[20]

Morcos, Benjamin. 2019. NengoFPGA. Master's thesis. MSc Thesis.

[21]

Mountcastle, Vernon B. 1997. The Columnar Organization of the Neocortex. Brain (1997).

[22]

Paszke et al. 2019. Pytorch: An Imperative style, High-Performance Deep Learning Library. In NeuroIPS'19.

Digital Library

[23]

Podobas et al. 2017. Designing and Cccelerating Spiking Neural Networks using OpenCL for FPGAs. In FPT'17.

[24]

Podobas et al. 2018. Hardware Implementation of POSITs and their application in FPGAs. In IPDPSW'18.

[25]

Ravichandran et al. 2020. Brain-like Approaches to Unsupervised Learning of Hidden Representations. arXiv preprint arXiv:2005.03476 (2020).

[26]

Ravichandran et al. 2020. Learning Representations in Bayesian Confidence Propagation Neural Networks. IJCNN'20 (2020).

[27]

Stathis et al. 2020. eBrainII: a 3 kW RealTime Custom 3D DRAM Integrated ASIC Implementation of a Biologically Plausible Model of a Human Scale Cortex. Journal of Signal Processing Systems (2020).

Cited By

Jahić BGuelfi NRies B(2023)SEMKIS-DSL: A Domain-Specific Language to Support Requirements Engineering of Datasets and Neural Network RecognitionInformation10.3390/info1404021314:4(213)Online publication date: 1-Apr-2023
https://doi.org/10.3390/info14040213
Giner-Miguelez JGómez ACabot J(2023)A domain-specific language for describing machine learning datasetsJournal of Computer Languages10.1016/j.cola.2023.10120976(101209)Online publication date: Aug-2023
https://doi.org/10.1016/j.cola.2023.101209

StreamBrain: An HPC Framework for Brain-like Neural Networks on CPUs, GPUs and FPGAs
1. Computer systems organization
  1. Architectures
    1. Other architectures
2. Computing methodologies
  1. Machine learning
    1. Machine learning approaches

Recommendations

An FPGA implementation for neural networks with the FDFM processor core approach

This paper presents a field programmable gate array FPGA implementation of a three-layer perceptron using the few DSP blocks and few block RAMs FDFM approach implemented in the Xilinx Virtex-6 family FPGA. In the FDFM approach, multiple processor cores ...
Representation Learning: A Review and New Perspectives

The success of machine learning algorithms generally depends on data representation, and we hypothesize that this is because different representations can entangle and hide more or less the different explanatory factors of variation behind the data. ...
Modeling and predicting performance of high performance computing applications on hardware accelerators

Hybrid-core systems speedup applications by offloading certain compute operations that can run faster on hardware accelerators. However, such systems require significant programming and porting effort to gain a performance benefit from the accelerators. ...

Comments

Please enable JavaScript to view thecomments powered by Disqus.

Information & Contributors

Information

Published In

cover image ACM Other conferences

HEART '21: Proceedings of the 11th International Symposium on Highly Efficient Accelerators and Reconfigurable Technologies

June 2021

76 pages

ISBN:9781450385497

DOI:10.1145/3468044

Copyright © 2021 ACM.

© 2021 Association for Computing Machinery. ACM acknowledges that this contribution was authored or co-authored by an employee, contractor or affiliate of the United States government. As such, the United States Government retains a nonexclusive, royalty-free right to publish or reproduce this article, or to allow others to do so, for Government purposes only.

In-Cooperation

German Research Foundation: German Research Foundation

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 21 June 2021

Permissions

Request permissions for this article.

Request Permissions

Check for updates

Author Tags

Qualifiers

Research-article
Research
Refereed limited

Funding Sources

European Commission

Conference

HEART '21

HEART '21: International Symposium on Highly Efficient Accelerators and Reconfigurable Technologies

June 21 - 23, 2021

Online, Germany

Acceptance Rates

Overall Acceptance Rate 22 of 50 submissions, 44%

Contributors

Other Metrics

View Article Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

2
Total Citations
View Citations
161
Total Downloads

Downloads (Last 12 months)24
Downloads (Last 6 weeks)2

Reflects downloads up to 12 Dec 2024

Other Metrics

View Author Metrics

Citations

Cited By

Jahić BGuelfi NRies B(2023)SEMKIS-DSL: A Domain-Specific Language to Support Requirements Engineering of Datasets and Neural Network RecognitionInformation10.3390/info1404021314:4(213)Online publication date: 1-Apr-2023
https://doi.org/10.3390/info14040213
Giner-Miguelez JGómez ACabot J(2023)A domain-specific language for describing machine learning datasetsJournal of Computer Languages10.1016/j.cola.2023.10120976(101209)Online publication date: Aug-2023
https://doi.org/10.1016/j.cola.2023.101209

View Options

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Full Access

Get this Publication

View options

PDF

View or Download as a PDF file.

eReader

View online with eReader.

Media

Figures

Other

Tables

View Table of Contents