Proceedings of the ACM International Conference on Computing Frontiers 2016
DOI: 10.1145/2903150.2903477
|View full text |Cite
|
Sign up to set email alerts
|

First impressions from detailed brain model simulations on a Xeon/Xeon-Phi node

Abstract: The development of physiologically plausible neuron models comes with increased complexity, which poses a challenge for many-core computing. In this work, we have chosen an extension of the demanding Hodgkin-Huxley model for the neurons of the Inferior Olivary Nucleus, an area of vital importance for motor skills. The computing fabric of choice is an Intel Xeon-Xeon Phi system, widely-used in modern computing infrastructure. The target application is parallelized with combinations of MPI and OpenMP. The best c… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
6
0

Year Published

2017
2017
2017
2017

Publication Types

Select...
2
1
1

Relationship

1
3

Authors

Journals

citations
Cited by 4 publications
(6 citation statements)
references
References 20 publications
0
6
0
Order By: Relevance
“…The Xeon Phi has also been used very successfully for bio-inspired neural networks, such as Convolutional Neural Networks for Deep Learning Systems [22]. On the other hand, similar difficulties to the GPUs in the acceleration of the complex HH models, are identified with Xeon Phi platforms even for less densely interconnected networks [2].…”
Section: Discussionmentioning
confidence: 99%
See 1 more Smart Citation
“…The Xeon Phi has also been used very successfully for bio-inspired neural networks, such as Convolutional Neural Networks for Deep Learning Systems [22]. On the other hand, similar difficulties to the GPUs in the acceleration of the complex HH models, are identified with Xeon Phi platforms even for less densely interconnected networks [2].…”
Section: Discussionmentioning
confidence: 99%
“…Traditional methods of computing, in which the common simulation tool-flows (such as MATLAB or specific neuromodeling tools like NEURON or Brian) are executed, are not up to the task of simulating neural networks of realistic sizes and high detail within a reasonable timeframe for brain research. High-Performance Computing (HPC) has been recently recognized as being able to provide a variety of solutions to cope with this limitation [2][3][4][5][6][7]. Unfortunately, the challenge of executing such simulation applications does not stop just at providing the necessary computational power.…”
Section: Introductionmentioning
confidence: 99%
“…In effect, the dataflow paradigm gradually degenerates to a sequential execution, making the application less scalable on the DFE. The Xeon Phi follows a similar trend, as the communication overhead between cores (which are interconnected through a moderately efficient ring topology [218]) increases, leading to similarly diminished scalability. Opposite to these accelerators, GPU scalability is largely improved.…”
Section: Resultsmentioning
confidence: 91%
“…. 4.2 Flowchart of the implementations discussed in this chapter [3] 16 The Knights Landing die organization [4]. Each tile consists of 2 cores that share an L2 cache.…”
Section: περίληψη στα ελληνικάmentioning
confidence: 99%
See 1 more Smart Citation