Spiking Neural Networks vs Artificial Neural Networks

Spiking Neural Networks

Neuromorphic computing based Spiking Neural Networks are so far the most extensive adaptation of brain inspired computing model. Unlike Artificial Neural Networks, Spiking Neural Networks receive inputs in form of spike and generate output in the form of spike. This is significant because the basic model of these neural networks resembles to the biological model of nervous system. These neural networks are based on neuromorphic computing chips. Unlike Artificial Neural Networks, in SNN the weight is called synaptic weight and it is adjusted by algorithm called spike time dependent plasticity (STDP).  

Spiking Neural Networks are the third generation of neural networks, the primary focus of Spiking Neural Networks is to narrow the gap between Machine learning and neuroscience using biologically realistic models of neuron.


Artificial Neural Network

Artificial neural networks are considered the foundation of Artificial intelligence. ANN are basically conventional neural networks made of small computational units called neurons. In today’s world, ANN is being embraced by almost every field including – Speech recognition, face recognition, power generation systems etc. Artificial Neural Networks can be modeled on conventional computers, which makes the implementation of Artificial Neural Networks relatively simple, whereas the SNN model requires a dedicated neuromorphic computing chip.


Neuromorphic computing        

The whole concept of neuromorphic computing chip is concerned with making physical computers more and more like human brain. Neuromorphic computing chips use the same physics of computation that is used by our own nervous system. Since our computer operate in binary, our codes and queries have to be structured in a rigid manner. Instead of using an electric signal to mean 1 or 0, the designers of these chips aim to make their neurons talk to each other like biological neurons do. To do this you need a precise electric current that flows between the synapse (space between neurons). This ability to transmit a gradient of understanding from neuron to neuron and have them all working together simultaneously means the neuromorphic chips could eventually be more energy efficient. Neuromorphic computing chip is a sophisticated and expensive gear, hence it has not yet made it’s way to the Mainstream. However, some of the universities and tech companies have succeeded in developing exploiting these chips. Following are some examples-

  1. SpiNNaker – University of Manchester
  2. TrueNorth-  IBM
  3. Loihi- Intel


Key difference between Spiking Neural Networks and Artificial Neural Networks




Wight adjustment methods

Spike time dependent plasticity (STDP)


Back propagation


Intended for research and understand the dynamics of human brain.


Numerous practical applications.


Require neuromorphic chip

Can be modeled on conventional computers


Overcomes von-neumann problem.

Doesn’t overcome von-neumann problem.

Leave A Reply

Ashpreet Kaur - Jul 2, 2021

Exercitation photo booth stumptown tote bag Banksy, elit small batch freegan sed. Craft beer elit seitan exercitation, photo booth et 8-bit kale chips proident chillwave deep v laborum. Aliquip veniam delectus, Marfa eiusmod Pinterest in do umami readymade swag.