Neuromorphic Computing Journey (Part 1)
Welcome to the blog series of Neuromorphic computing. In this series we will look into hardware and software aspects of Neuromorphic Computing and spiking neural networks.
Lets start with what is Neuromorphic Computing and why all of the sudden researches are getting attracted towards it? So far, many computing techniques are implemented for different general/ specific applications but energy efficiency is still a major issue in every computation. The more complex computation it is the more energy our system need to compute it. With a boom in more complex systems, researchers are very focused to develop a systems that can perform complex computation with less energy consumption.
Many researchers in past tried to draw motivation from how the human brain performs. Efforts were made to understand the complexity of human brain and how it compute so much information with so less energy in short amount of time. Human brain uses only 20 watt and this amount of energy it can do wonders, which we see in daily life. Computer technology is still far behind. Though the recent development in area of AI opens many door to make computer do predictions but there is still lot to do in this area.
In this blog we will start from the history of Neuromorphic Computing. Most of the information in this blog is taken from the Steve Furber’s SpiNNAker book. (what is SpiNNaker? be patient, we will talk about is latter in detail)
Beginning of Neuromorphic Computing
Many people were inspired by human brain in history. It is impossible to discuss everyone of them here. Here I am presenting one example of Ada Lovelace who is one of the pioneer in this field. Some of lines from her notes are as follows:
‘I have my hopes, and very distinct ones too, of one day getting cerebral phenomena such that I can put them into mathematical equations — in short, a la or laws for the mutual actions of the molecules of brain. I hope to bequeath to the generations a calculus of the nervous system.’
Her partner was Charles Babbage and together with their team they designed very early mechanical computing system and Ada thought about such algorithm which such computers might able to learn. Some people believe that she was wold first computer programmer. No wonder Ada was far way ambitious in her thoughts because now even after almost 200 years ago we are still no way near to implement human brain into mathematical equations.
Another example in this field is of famous Alan Turing. Alan moved to Manchester, UK to work on the 1930’s universal computing machine. During his time in Manchester, he worked on multiple topic but the one interesting here is his paper with title ‘Computing Machinery and Intelligence’. This paper was focusing on ‘can computer think?’. It was only 2 years into the modern computer and Turing was already started thinking that how far this technology can go. Turing thought that by the end of 20th Century, computers will have enough memory to pass the tests he designed, now which is known as Turing tests.
The Turing was right about the computer capabilities but according to Steve Furber, famous computer scientist of 21st century, the reason we are still unable to make machines think like human brain is because still don’t have complete understanding of how our brain works and how it processes data.
In 1980s, Carver Mead worked at Caltech and began the field which now is known as Neuromorphic computing. His interest was to understand the analogy between ion channels in neurons and flow of electrons in the sub-threshold regions of transistors. He proposed the idea of creating analogue electronic sensors and also created such chips. To this day, research is going on in this area.
That’s all for now. In the next section of this series we will discuss about the Human Brain Project (HBP) which is a flagship project of European Union and its main focus is to develop a brain like devices.
Thank you for reading this blog. Don’t forget to follow.
