A Resource Allocation Method for Neuron Computer Operating System

Wang Fengjuan; Lü Pan; Jin Ouwen; Xing Qinghui; Deng Shuiguang

doi:10.7544/issn1000-1239.202330422

Journal of Computer Research and Development > 2023 > 60(9): 1948-1959. > DOI: 10.7544/issn1000-1239.202330422

Wang Fengjuan, Lü Pan, Jin Ouwen, Xing Qinghui, Deng Shuiguang. A Resource Allocation Method for Neuron Computer Operating System[J]. Journal of Computer Research and Development, 2023, 60(9): 1948-1959. DOI: 10.7544/issn1000-1239.202330422

Citation:

PDF (2263 KB)

A Resource Allocation Method for Neuron Computer Operating System

1.
Zhejiang Lab, Hangzhou 311121
2.
College of Computer Science and Technology, Zhejiang University, Hangzhou 310058

Funds: This work was supported by the National Key Research and Development Program of China (2022YFB4500100).

More Information

Author Bio:
Wang Fengjuan: born in 1987. Master. Her main research interests include computer architecture and brain-inspired computing OS

Lü Pan: born in 1981. PhD candidate. His main research interests include computer architecture and brain-inspired computing system software

Jin Ouwen: born in 1998. PhD candidate. His main research interests include computer architecture and brain-inspired computing

Xing Qinghui: born in 1999. PhD candidate. His main research interests include computer architecture and brain-spired computing

Deng Shuiguang: born in 1979. PhD, professor, PhD supervisor. Distinguished member of CCF. His main research interests include service computing, edge computing, software engineering, and big data
Received Date: May 30, 2023
Revised Date: July 22, 2023
Available Online: August 01, 2023

Graphical Abstract

Abstract

Abstract

Neuromorphic hardware is a specialized computer system designed for running spiking neural network (SNN) applications. With the increasing scale of hardware resources and the challenge of concurrent execution of numerous SNN applications, efficiently allocating neuromorphic hardware resources to SNN applications has become highly challenging. We propose a resource allocation process for a neural computer operating system that maximizes the decoupling of resource allocation from compiler. We allocate hardware resources and corresponding input-output routing for SNN applications only when loading them onto the neuromorphic hardware. Furthermore, we introduce the innovative maximum empty rectangle (MER) algorithm to address the management and dynamic allocation of neuromorphic hardware resources. Additionally, we present a resource allocation algorithm that minimizes the communication cost of spiking-based input-output in SNNs, aiming to reduce energy consumption, latency, and resource fragmentation. Experimental results demonstrate that our algorithm outperforms existing approaches in terms of energy consumption, latency, and fragmentation rate.
- spiking neural network,
- neuromorphic computing hardware,
- spiking input output,
- resource allocation,
- maximum empty rectangle (MER)

FullText(HTML)

References (29)

References

[1]	Waldrop M M. The chips are down for Moore’s law[J]. Nature, 2016, 530(7589): 144−147 doi: 10.1038/530144a
[2]	Zhang Bo, Shi Luping, Song Sen. Creating more intelligent robots through brain-inspired computing[J]. Science, 2016, 354: 4−9
[3]	Maass W. Networks of spiking neurons: The third generation of neural network models[J]. Neural Networks, 1997, 10(9): 1659−1671 doi: 10.1016/S0893-6080(97)00011-7
[4]	Roy K, Jaiswal A, Panda P. Towards spike-based machine intelligence with neuromorphic computing[J]. Nature, 2019, 575(7784): 607−617 doi: 10.1038/s41586-019-1677-2
[5]	Furber S B, Galluppi F, Temple S, et al. The spiNNaker project[J]. IEEE, 2014, 102(5): 652−665 doi: 10.1109/JPROC.2014.2304638
[6]	Scholze S, Eisenreich H, Höppner S, et al. A 32 GBit/s communication SoC for a waferscale neuromorphic system[J]. Integration, 2012, 45(1): 61−75 doi: 10.1016/j.vlsi.2011.05.003
[7]	Merolla P A, Arthur J V, Alvarez-icaza R, et al. A million spiking-neuron integrated circuit with a scalable communication network and interface[J]. Science, 2014, 345(6197): 668−673 doi: 10.1126/science.1254642
[8]	Davies M, Srinivasa N, Lin T H, et al. Loihi: A neuromorphic manycore processor with on-chip learning[J]. IEEE Micro, 2018, 38(1): 82−99 doi: 10.1109/MM.2018.112130359
[9]	Pei Jing, Deng Lei, Song Sen, et al. Towards artificial general intelligence with hybrid Tianjic chip architecture[J]. Nature, 2019, 572: 106−111 doi: 10.1038/s41586-019-1424-8
[10]	Shi Luping, Pei Jing, Deng Ning, et al. Development of a neuromorphic computing system[C]//Proc of 2015 IEEE Int Electron Devices Meeting (IEDM). Piscataway, NJ: IEEE, 2015: 72−75
[11]	Moradi S, Qiao Ning, Stefanini F, et al. A scalable multicore architecture with heterogeneous memory structures for dynamic neuromorphic asynchronous processors (DYNAPs)[J]. IEEE Transactions on Biomedical Circuits and Systems, 2017, 12(1): 106−122
[12]	Benjamin B V, Gao Peiran, McQuinn E, et al. Neurogrid: A mixed-analog-digital multichip system for large-scale neural simulations[J]. Proceedings of the IEEE, 2014, 102(5): 699−716 doi: 10.1109/JPROC.2014.2313565
[13]	Ma De, Shen Juncheng, Gu Zonghua, et al. Darwin: A neuromorphic hardware co-processor based on spiking neural networks[J]. Systems Architecture, 2017, 77: 43−51 doi: 10.1016/j.sysarc.2017.01.003
[14]	Galluppi F, Davies S, Rast A, et al. A hierachical configuration system for a massively parallel neural hardware platform[C]//Proc of the 9th Conf on Computing Frontiers. New York: Association for Computing Machinery, 2012: 183−192
[15]	Balaji A, Adiraju P, Kashyap H J, et al. PyCARL: A PyNN interface for hardware-software co-simulation of spiking neural network[C]//Proc of Int Joint Conf on Neural Networks (IJCNN). Piscataway, NJ: IEEE, 2020: 19−24
[16]	Balaji A, Das A, Wu Yuefeng, et al. Mapping spiking neural networks to neuromorphic hardware[J]. IEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2020, 28(1): 76−86 doi: 10.1109/TVLSI.2019.2951493
[17]	Song Shihao, Chong Harry, Balaji A, et al. DFSynthesizer: Dataflow-based synthesis of spiking neural networks to neuromorphic hardware[J]. ACM Transactions on Embedded Computing Systems, 2021, 21(3): 1−35
[18]	Jin Ouwen, Xing Qinghui, Li Ying, et al. Mapping very large scale spiking neuron network to neuromorphic hardware[C]//Proc of the 28th ACM Int Conf on Architectural Support for Programming Languages and Operating Systems. New York: ACM, 2023: 419−432
[19]	Müller E, Schmitt S, Mauch C, et al. The operating system of the neuromorphic BrainScaleS-1 system[J]. Neurocomputing, 2022, 501: 790−810 doi: 10.1016/j.neucom.2022.05.081
[20]	Rowley A G D, Brenninkmeijer C, Davidson S, et al. SpiNNTools: The execution engine for the SpiNNaker platform[J]. Frontiers Neurosci, 2019, 13: 231−231 doi: 10.3389/fnins.2019.00231
[21]	Lin C-K, Wild A, Chinya G N, et al. Programming spiking neural networks on Intel’s Loihi[J]. Computer, 2018, 51(3): 52−61 doi: 10.1109/MC.2018.157113521
[22]	Bazargan K, Kastner R, Sarrafzadeh M. Fast template placement for reconfigurable computing systems[J]. IEEE Design and Test of Computers, 2000, 17: 68−83 doi: 10.1109/54.825678
[23]	Handa M, Vemuri R. An efficient algorithm for finding empty space for online FPGA placement[C]//Proc of the 41st Design Automation Conf. Piscataway, NJ: IEEE, 2004: 960−965
[24]	Galluppi F, Davies S, Rast A, et al. A hierachical configuration system for a massively parallel neural hardware platform[C]//Proc of the 9th Conf on Computing Frontiers. New York: ACM, 2012: 183−192
[25]	Heathcote J. Building and operating large-scale SpiNNaker machines[D]. Manchester, UK: The University of Manchester, 2016
[26]	Sawada J, Akopyan F, Cassidy A S, et al. TrueNorth ecosystem for brain-inspired computing: Scalable systems, software, and applications[C]// Proc of the Int Conf for High Performance Computing, Networking, Storage and Analysis(SC’16) . Piscataway, NJ: IEEE, 2016: 130−141
[27]	Deng Shuiguang, Lv Pan, Jin Ouwen, et al. Darwin-S: A reference software architecture for brain-inspired computers[J]. Computer, 2022, 55(5): 51−63 doi: 10.1109/MC.2022.3144397
[28]	Wang Guohua, Liu Song, Nie Jing, et al. An online task placement algorithm based on maximum empty rectangles in dynamic partial reconfigurable systems[C]//Proc of 2017 NASA/ESA Conf on Adaptive Hardware and Systems (AHS). Piscataway, NJ: IEEE, 2017: 180−185
[29]	Jylanki J. A thousand ways to pack the bin−a practical approach to two-dimensional rectangle bin packing[J/OL]. [2023-01-18]http://clb. demon. fi/files/RectangleBinPack.pdf, 2010