The opportunities of neural network
We can think of the internet as the giant neural network. And technically we can connect every computer and mobile device that is using internet-connection to the same neural network. The thing is if we would someday make that kind of action, we could solve many problems like is it possible that Riemann's conjecture has the zero-point.
The search for alien life has brought new ways to think about life, and the second thing is that research has been giving a push to innovative solutions of neural-network-based computing. In the neural network, the computing happens in the smaller data-handling units, which are connected. If we would use the neural-based networks for analyzing data, the system that is needed can be consist of normal personal computers. And in this model, there is no need for the giant servers or supercomputers, the neural network means that every single unit in this type of system is handling their part of the operation.
That means that even the single data-handling unit in the neural network is not very powerful, those computers are working together. In this system, all workstations that are connected to some network are acting together for finding out the solution for the problem. And in the neural network, the system can benefit the "empty time" of the microprocessors for making the operations more effective.
The thing, why the neural-networks are so interesting, is that if we would connect the workstations at some company together that they are working as the supercomputer, that gives the outstanding data handling ability to that corporation. This method has been created for making the calculations for the SETI program lighter and faster. But it is also used in the Render farm and Blender (1), the open-source animation tool. And by using the neural network the autonomous drone swarms can operate independently.
This technology allows connecting the different types of systems, and that gives the neural networks many useful targets from the rescue to the military and space science. The autonomously operating drone-swarms can be dropped to the atmospheres of the Venus and Titan moon, as well as the atmospheres to giant gas planets. Those drone swarms can be operated in large areas and the loose of one part of the swarm doesn't risk the entire missions.
Drone swarms can search the enormous areas, and that's why the military and rescue services are interesting about those systems. But the thing is that neural networks allow also regular schools and companies can create graphical and other types of solutions, which are normally possible to make by using the supercomputer.
The neural network can allow the architects the possibility to make flow stimulations by using normal-price computers. The thing is that the neural network can be more than we ever thought. That means that there is a theoretical possibility that all computers on the Earth would be connected to the one neural network, which gives outstanding opportunities to make simulations and calculations.
But the reason, why this thing is impossible is that the political resistance for that is high. And that's why this kind of idea is pure theory. Technically that can be done, but the laws are prohibiting that. Also, two neural networks can be linked together or the supercomputers can connect to the neural networks, which gives them very high-level and powerful computing capacity.
Comments
Post a Comment