A novel CPU/GPU simulation environment for large-scale biologically realistic neural modeling

Hoang, Roger V.; Tanna, Devyani; Bray, Laurence C. Jayet; Dascalu, Sergiu M.; Harris, Frederick C.

doi:10.3389/fninf.2013.00019

Cited by 52 publications

(46 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is particularly true in the case of GPU simulations, as it is difficult to make maximally efficient use of GPU resources. Consequently, almost all GPU-based simulators for spiking neural networks have not made it possible to easily create new user-defined neuron models [4][5][6][7][8] . The exceptions are GeNN, the package Brian2CUDA 13 currently under development, and ANNarchy 14 , which is discussed below.…”

Section: Discussionmentioning

confidence: 99%

Brian2GeNN: accelerating spiking neural network simulations with graphics hardware

Stimberg

Goodman

Nowotny

2020

Sci Rep

View full text Add to dashboard Cite

is a popular Python-based simulator for spiking neural networks, commonly used in computational neuroscience. GeNN is a C++-based meta-compiler for accelerating spiking neural network simulations using consumer or high performance grade graphics processing units (GPUs). Here we introduce a new software package, Brian2GeNN, that connects the two systems so that users can make use of GeNN GPU acceleration when developing their models in Brian, without requiring any technical knowledge about GPUs, C++ or GeNN. The new Brian2GeNN software uses a pipeline of code generation to translate Brian scripts into C++ code that can be used as input to GeNN, and subsequently can be run on suitable NVIDIA GPU accelerators. From the user's perspective, the entire pipeline is invoked by adding two simple lines to their Brian scripts. We have shown that using Brian2GeNN, two non-trivial models from the literature can run tens to hundreds of times faster than on CPU. GPU acceleration emerged when creative academics discovered that modern graphics processing units (GPUs) could be used to execute general purpose algorithms, e.g. for neural network simulations 1,2. The real revolution occurred when NVIDIA corporation embraced the idea of GPUs as general purpose computing accelerators and developed the CUDA application programming interface 3 in 2006. Since then, GPU acceleration has become a major factor in high performance computing and has fueled much of the recent renaissance in artificial intelligence. One of the remaining challenges when using GPU acceleration is the high degree of insight into GPU computing architecture and careful optimizations needed in order to achieve good acceleration, in spite of the abstractions that CUDA offers. A number of simulators have used GPUs to accelerate spiking neural network simulations, but the majority do not allow for easily defining new models, relying instead on a fixed set of existing models 4-8. Since 2010 we have been developing the GPU enhanced neuronal networks (GeNN) framework 9 that uses code generation techniques 10,11 to simplify the use of GPU accelerators for the simulation of spiking neural networks. GPUs, and in particular GeNN, have been shown to enable efficient simulations compared to CPUs and even compared to dedicated neuromorphic hardware 12. Other simulators that have taken this code generation approach are Brian2CUDA 13 (currently under development) and ANNarchy 14 (Linux only). Brian is a general purpose simulator for spiking neural networks written in Python, with the aim of simplifying the process of developing models 15-17. Version 2 of Brian 18 introduced a code generation framework 10,19 to allow for higher performance than was possible in pure Python. The design separates the Brian front-end (written in Python) from the back-end computational engine (multiple possibilities in different languages, including C++), and allows for the development of third party packages to add new back-ends. Here, we introduce the Brian2GeNN software interface we have developed to a...

show abstract

Section: Discussionmentioning

confidence: 99%

Brian2GeNN: accelerating spiking neural network simulations with graphics hardware

Stimberg

Goodman

Nowotny

2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…This is particularly true in the case of GPU simulations, as it is difficult to make maximally efficient use of GPU resources. Consequently, almost all GPU-based simulators for spiking neural networks have not made it possible to easily create new user-defined neuron models (Nageswaran et al, 2009;Fidjeland and Shanahan, 2010;Mutch et al, 2010;Hoang et al, 2013;Bekolay et al, 2014). The exceptions are GeNN, the package Brian2CUDA (Augustin et al, 2018) currently under development, and ANNarchy (Vitay et al, 2015), which is discussed below.…”

Section: Discussionmentioning

confidence: 99%

Brian2GeNN: a system for accelerating a large variety of spiking neural networks with graphics hardware

Stimberg

Goodman

Nowotny

2018

Preprint

View full text Add to dashboard Cite

Brian" is a popular Python-based simulator for spiking neural networks, commonly used in computational neuroscience. GeNN is a C++-based meta-compiler for accelerating spiking neural network simulations using consumer or high performance grade graphics processing units (GPUs). Here we introduce a new software package, Brian2GeNN, that connects the two systems so that users can make use of GeNN GPU acceleration when developing their models in Brian, without requiring any technical knowledge about GPUs, C++ or GeNN. The new Brian2GeNN software uses a pipeline of code generation to translate Brian scripts into C++ code that can be used as input to GeNN, and subsequently can be run on suitable NVIDIA GPU accelerators. From the user's perspective, the entire pipeline is invoked by adding two simple lines to their Brian scripts. We have shown that using Brian2GeNN, typical models can run tens to hundreds of times faster than on CPU. different languages, including C++), and allows for the development of third party packages to add new back-ends.Here, we introduce the Brian2GeNN software interface we have developed to allow running Brian models on a GPU via GeNN. We analysed the performance for some typical models and find that -depending on the CPU and GPU used -performance can be tens to hundreds of times faster. ResultsWe benchmarked Brian2GeNN on two model networks we will name "COBAHH" and "Mbody" in the remainder of this paper. COBAHH is an implementation of a benchmark network described in Brette et al. (2007). Essentially, this benchmark model consists of N Hodgkin-Huxley-type neurons, modified from the model in Traub and Miles (1991), 80% of which form excitatory synapses and 20% inhibitory synapses. All neurons were connected to all other neurons randomly with a connection probability chosen such that each neuron received on average 1000 connections for large models, or connections from all other neurons if the number of neurons was less than 1000.Mbody is an implementation of the mushroom body model of Nowotny et al. (2005) but unlike in the original publication also with a similar neuron model to the one used for the COBAHH benchmark. The model was used with 100 input neurons, 100 output neurons and varying numbers N of Kenyon cells (hidden layer). Projection neurons in the input layer are connected with fixed probability of 15% to Kenyon cells. Up to N = 10, 000 Kenyon cells they are connected all-to-all to the output neurons, and for N > 10, 000, they are connected randomly with probability chosen such that the output neurons receive input from on average 10, 000 Kenyon cells.Both models were integrated with an exponential Euler algorithm at 0.1ms time steps. The benchmarks presented here were obtained using the GeNN sparse matrix representation for synaptic connections.We benchmarked the models on different systems and with different backends. The GeNN backend through the Brian2GeNN interface presented here was compared to the "C++ standalone" backend included with the Brian simulator which runs on the CPU...

show abstract

“…The BRIAN Computation Laboratory of the Department of Computer Science and Engineering of the University of Nevada presented an updated version of their NeoCortical Simulator, an open-source CPU/GPU simulation environment for large-scale biological networks [6]. The simulator is currently able to simulate 1M neurons and 100M synapses in quasi real time using eight nodes, each having 2 video cards.…”

Section: Related Workmentioning

confidence: 99%

GPU Implementation of Neural-Network Simulations Based on Adaptive-Exponential Models

Neofytou

Chatzikonstantis

Magkanaris

et al. 2019

2019 IEEE 19th International Conference on Bioinformatics and Bioengineering (BIBE)

View full text Add to dashboard Cite

Detailed brain modeling has been presenting significant challenges to the world of high-performance computing (HPC), posing computational problems that can benefit from modern hardware-acceleration technologies. We explore the capacity of GPUs for simulating large-scale neuronal networks based on the Adaptive Exponential neuron-model, which is widely used in the neuroscientific community. Our GPU-powered simulator acts as a benchmark to evaluate the strengths and limitations of modern GPUs, as well as to explore their scaling properties when simulating large neural networks. This work presents an optimized GPU implementation that outperforms a reference multicore implementation by 50×, whereas utilizing a dual-GPU configuration can deliver a speedup of 90× for networks of 20, 000 fully interconnected AdEx neurons.

show abstract

A novel CPU/GPU simulation environment for large-scale biologically realistic neural modeling

Cited by 52 publications

References 26 publications

Brian2GeNN: accelerating spiking neural network simulations with graphics hardware

Brian2GeNN: accelerating spiking neural network simulations with graphics hardware

Brian2GeNN: a system for accelerating a large variety of spiking neural networks with graphics hardware

GPU Implementation of Neural-Network Simulations Based on Adaptive-Exponential Models

Contact Info

Product

Resources

About