Search: onr:"swepub:oai:DiVA.org:kth-310412" >
StreamBrain :
StreamBrain : An HPC Framework for Brain-like Neural Networks on CPUs, GPUs and FPGAs
-
- Podobas, Artur (author)
- KTH,Beräkningsvetenskap och beräkningsteknik (CST)
-
- Svedin, Martin (author)
- KTH,Beräkningsvetenskap och beräkningsteknik (CST)
-
- Chien, Wei Der (author)
- KTH,Beräkningsvetenskap och beräkningsteknik (CST)
-
show more...
-
- Peng, Ivy B. (author)
- Lawrence Livermore National Laboratory, CA, USA
-
- Ravichandran, Naresh Balaji (author)
- KTH,Beräkningsvetenskap och beräkningsteknik (CST)
-
- Herman, Pawel, 1979- (author)
- KTH,Beräkningsvetenskap och beräkningsteknik (CST)
-
- Lansner, Anders, Professor, 1949- (author)
- KTH,Beräkningsvetenskap och beräkningsteknik (CST),Stockholm University, Stockholm, Sweden
-
- Markidis, Stefano (author)
- KTH,Beräkningsvetenskap och beräkningsteknik (CST)
-
show less...
-
(creator_code:org_t)
- 2021-06-21
- 2021
- English.
-
In: ACM International Conference Proceeding Series. - New York, NY, USA : Association for Computing Machinery (ACM).
- Related links:
-
http://arxiv.org/pdf...
-
show more...
-
https://urn.kb.se/re...
-
https://doi.org/10.1...
-
show less...
Abstract
Subject headings
Close
- 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.
Subject headings
- NATURVETENSKAP -- Data- och informationsvetenskap -- Datavetenskap (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences -- Computer Sciences (hsv//eng)
Keyword
- AI
- BCPNN
- Emerging Machine Learning
- FPGA
- GPU
- HPC
- Neural networks
- Representation learning
- Unsupervised learning
- Backpropagation
- Deep learning
- Digital arithmetic
- Field programmable gate arrays (FPGA)
- Learning systems
- Problem oriented languages
- Program processors
- Application area
- Benchmark datasets
- Domain specific languages
- Floating points
- High performance computing systems
- Learning methods
- Multiple domains
- Theoretical foundations
Publication and Content Type
- ref (subject category)
- kon (subject category)
To the university's database
- By the author/editor
-
Podobas, Artur
-
Svedin, Martin
-
Chien, Wei Der
-
Peng, Ivy B.
-
Ravichandran, Na ...
-
Herman, Pawel, 1 ...
-
show more...
-
Lansner, Anders, ...
-
Markidis, Stefan ...
-
show less...
- About the subject
-
- NATURAL SCIENCES
-
NATURAL SCIENCES
-
and Computer and Inf ...
-
and Computer Science ...
- Articles in the publication
- ACM Internationa ...
- By the university
-
Royal Institute of Technology