A Mixed-Signal Structured AdEx Neuron for Accelerated Neuromorphic Cores

被引:39
作者
Aamir, Syed Ahmed [1 ]
Mueller, Paul [1 ]
Kiene, Gerd [1 ]
Kriener, Laura [1 ]
Stradmann, Yannik [1 ]
Gruebl, Andreas [1 ]
Schemmel, Johannes [1 ]
Meier, Karlheinz [1 ]
机构
[1] Heidelberg Univ, Kirchhoff Inst Phys, D-69120 Heidelberg, Germany
基金
欧盟地平线“2020”;
关键词
65 nm CMOS; Analog integrated circuits; adaptation; AdEx; bursting; dendrites; exponential; integrate-and-fire; leaky; LIF; multi-compartment; neuromorphic; neuron; NMDA; spiking; SPIKING NEURONS; MODEL; SYNAPSES; CIRCUIT; ARRAY; BACKPROPAGATION; DENDRITES; DIRECTION; DYNAMICS; CELLS;
D O I
10.1109/TBCAS.2018.2848203
中图分类号
R318 [生物医学工程];
学科分类号
0831 ;
摘要
Here, we describe a multicompartment neuron circuit based on the adaptive-exponential I&F (AdEx) model, developed for the second-generation BrainScaleS hardware. Based on an existing modular leaky integrate-and-fire (LIF) architecture designed in 65-nm CMOS, the circuit features exponential spike generation, neuronal adaptation, intercompartmental connections as well as a conductance-based reset. The design reproduces a diverse set of firing patterns observed in cortical pyramidal neurons. Further, it enables the emulation of sodium and calcium spikes, as well as N-methyl-D-aspartate plateau potentials known from apical and thin dendrites. We characterize the AdEx circuit extensions and exemplify how the interplay between passive and nonlinear active signal processing enhances the computational capabilities of single (but structured) on-chip neurons.
引用
收藏
页码:1027 / 1037
页数:11
相关论文
共 76 条
[1]  
Aamir SA, 2016, PROC EUR SOLID-STATE, P71
[2]  
[Anonymous], THESIS
[3]  
[Anonymous], 2018, ACCELERATED LIF NEUR
[4]  
[Anonymous], FRONTIERS NEUROSCIEN
[5]  
[Anonymous], 2017, IEEE BIOM CIRC SYST
[6]  
[Anonymous], BRAINSCALES 2 NOVEL
[7]  
[Anonymous], 2011, Custom Integrated Circuits Conference
[8]  
[Anonymous], THESIS
[9]  
[Anonymous], 2017, THESIS
[10]  
[Anonymous], THESIS