Introduction to this blog

Consciousness can be generated by the process of computation! This is the claim of neuroscientists and computer scientists. However, the Vedic literature provides a clear idea that conscious by its nature is not a matter and hence it cannot be generated by any kind of complex computation that is due to matter. This blog is to explore the science of consciousness from Vedic perspective.

Saturday, March 13, 2021

Computational Model and Human Brain

Computation has become so much important that it has become impossible to live without it in modern times. The computation has spread across our day-to-day life in various ways. We use cell phone, we use various other computational devices for our livelihood in every home utilities as well as in many other technology. Further, computation has become a tool for the advancement of science in the pursuance of knowledge. It is of course true that we are yet to define what we suppose to be calling as a knowledge.  We have discussed on the definition of knowledge in other article. In the set of knowledge that has been learned so far by the scholars, the computation is one such which needs to be understood further. The curiosity has lead the investigation further to learn about the nature of computation and computation in  the Nature that we live in. This article is a set of thoughts to understand the nature of the human brain in the light of computational model. Our interest is to understand the structure and functionality of the human brain in contrast with the computational model. Is computation one of the fundamental principles of Nature? Does this material nature also performs a unique kind of computation in order to sustain the activities? As computer science has evolved theoretically as well as physically with the advent of different technology and computational model, it is appropriate to contrast the current computational technology with that of the human brain. 

The computation has been defined in terms of an abstract machine called Turing machine by Alan Turing a British mathematician. This definition has become a model for the theoretical study of computer science. The model describes the process of computation as information processing where the information is represented in the form of a set of well-defined symbols. The processing is controlled by an automated machine which can be called a controller, technically this can be termed as a state machine. Further, the processing requires an ability to remember the past for which a memory device is used called tap. With these three components, the Turing machine is constructed which defines what computation suppose to be. Most of the modern computational devices which are most popular are based on this definition. The physical counterpart of this definition is the Von-Neumann architecture (of course there is another architecture called Stanford Architecture conceptually similar with VNA.) which describes further how a machine could process the symbols in a step-by-step manner to reach at a conclusive result. Therefore, it can be said that the Von-Neumann architecture introduces additional components with data and control paths to facilitate the flow of information and data. This architecture, VNA, has been widespread now in almost all the processors that are currently being manufactured. Among many such processors, the general-purpose processors are most popular due to its basic properties of being programmable. The programmability feature implies that the same processor could be used to carry out multiple tasks. It is for that reason called as a general-purpose processor. It is the program that is unique for a particular task. The underlying processor is common. At the first look when we contrast this architecture with that of our human brain we find that our brain is something different from the architecture that we have just described. As time progress and along with that the human needs which can be termed as applications also is emerging with various demand. One such demand is performance. The applications that the modern computer process are complex in nature and therefore they demand high-performance computing. As the demand for high-performance computing rises, there have been several pieces of research that attempt to meet this performance demand. One of the interesting advancements in this direction to note is the development of application-specific accelerators. Here comes the something worth noting point which is the subject of this article. If we carefully observe the application-specific accelerators, it is one-to-one with the architecture of the human brain. Whereas not the general-purpose processor architecture. 

The interesting part to investigate is how far the application-specific accelerator  architecture would be correlated with the human brain architecture. And, would it be the appropriate architecture for neural processing? 

We will investigate further here.  

Reference: 1. Image is taken from:

Jaynarayan, Bangalore, India

No comments: