International Journal of Scientific Research and Review ISSN NO: 2279-543X

Review of Technologies

K. Nivethithaan Terrin J Mario S. Prasath K. Balaji Gokul R UG Scholar B.E UG Scholar B.E UG Scholar B.E UG Scholar B.E Pereira Mechatronics Mechatronics Mechatronics Mechatronics UG Scholar B.E Engineering Engineering Engineering Engineering Mechatronics SNS College of SNS College of SNS College of SNS College of Engineering Technology Technology Technology Technology SNS College of Coimbatore, Coimbatore, Coimbatore, Coimbatore, Technology Tamil Nadu. Tamil Nadu. Tamil Nadu. Tamil Nadu. Coimbatore,

Tamil Nadu.

Abstract — Traditionally computers are based on semi- Optical are commonly used in optic fiber cables conductors namely silicon or germanium. These types of to transmit information to long distances. But optical compilers depend on the passage of electrical current through computing goes way beyond just transmission of data. made of small semi-conductors. Optical Computers Some of the hypothesized advantages of Optical are similar to traditional semi-conductor based digital computers Computers over Traditional Computers include: except that rather than using electrical impulses, these computers use of for computation. Although this technology is i. The switching time between the 2 states in an optical still in its infancy, there are a lot of methods which can be used is much faster than in a for this purpose. This paper receives the various methods used in transistor. optical computing and plots a future path for development of ii. They can be directly linked to existing optic fiber optical computing. cables, whereas traditional systems require photo Keywords—component; formatting; style; styling; insert (key and receptors which caused extra complexity words) and creates delay in the system. iii. Controversially, optical transistors can also reduce the energy required for switching between the INTRODUCTION :- various states. This argument is questionable as it is Digital Computing refers to the use of just 2 values not proven beyond doubt that they can be used for ‘ON’ and ‘OFF’ for computation. Traditional, Digital large scale computation. computers are of semi-conductors, and they rely on electrical iv. Finally, it has been argued that using optical impulses for computation. A typical truth table for a semi- transistors can reduce the energy required to run the conductor based digital computer is computer. This can enable a system to run on lower A B X&Y X|Y X! Y! supply requirements. 0 0 0 0 1 1 Although optical computers have so many advantages, 0 1 0 1 1 0 there are certain troubling issues that can hinder this 1 0 0 1 0 1 development. Some of such issues include: 1 1 1 1 0 0

i. As computation is a non- linear process, and multiple But, in reality there are multiple ways to build a digital must interact in order for an accurate answer. computer than using electricity such as using reversible chemical reactions, using Potassium ions, and most notably Light as it is an EM wave, can only interact with using light signals. Such computers are very similar to the other EM waves in presence of . This traditional digital computers as they too use binary logic for strength of interaction between EM waves is weaker computation. This means that a processes meant for traditional than that of electronic signals in a conventional digital computers can work well with optical computers. computer. This means that a fully optical computer The biggest difference between optical and digital would require more power and the transistor must be computers is that, unlike traditional computers optical of larger dimension than in traditional computer. computers use optical transistors rather than semi-conductor ii. Light waves, as with all electromagnetic waves are based transistors bound by the transforms limit, meaning that for every pulse of the wave there must be a time limit attached

Volume 7, Issue 3, 2018 37 http://dynamicpublisher.org/ International Journal of Scientific Research and Review ISSN NO: 2279-543X

to it. This problem becomes magnified when the light constricted using a super-cooled gas. When the light hits one is subjected to depression in a medium and the of the electrons of the gas, the gas becomes charged due to the spectral bandwidth are taken into account. Correcting absorption of energy and the mirror becomes opaque. this would require a transmission of multiple ultra- short pulses over various frequencies and bandwidth. TIME DELAY OPTICAL COMPUTING :- iii. They can be used only for such application where many paradigms of Optical Computing are focused The basic idea behind Time Delay Optical Computing on solving on a fixed set of problems such as NR is to defer light (or some other flag) keeping in mind the end goal to perform valuable computations. Of intrigue is tackle complete problems or Travelling Salesman problem. NP-Complete issues as those are extremely difficult to solve Further research is required to fully understand the using Conventional PCs. scope and application of fully Optical Computers. There are 2 fundamental properties of light that can be utilized as a part of this approach: LOGIC :- 1. The light can be deferred by going it through an Photonic logic refers to the logic that can be applied to optical fiber of a specific length. optical computers. This type of logic is extremely similar to 2. The light can be part into numerous (sub) rays. This conventional binary logic in which ‘O’ represents no light property is likewise basic since we can assess pulse and ‘|’ represents a light pulse. numerous arrangements in a similar time. The biggest advantage of such logic is that unlike the other approaches which either rely on wavelength or non- When tackling an issue with time-defers the following steps standard behaviors of light. This method is extremely easy to must be taken: execute and is highly compatible to existing computer 1. Initial step is to create a chart or graph like structure technology and like traditional binary logic, the logic gates produced using optical links and splitters. Each graph [AND, OR, NOT, XOR, XNOR, NAND, NOR] can be used to has a Start hub and a Destination hub. perform logical operations. 2. The light enters through the Start hub and navigates the chart until the point that it achieves the CONSTRUCTION OF OPTICAL TRANSISTORS :- Destination. It is postponed when going through An , also called an optical . It is curves and isolated inside hubs. a device that can be used to switch or amplify light signals. It 3. The light is stamped when going through a circular is highly similar to an electronic transistor, except in an segment or through a hub so we can without much of electronic transistor electrons take the place of light signals. a stretch distinguish that reality at the Destination An Optical transistor must be constructed from non- linear media. Non-Linear media are materials in which the hub. dielectric polarization P (polarization occurring in an insulating medium) responds non linearly (i.e. the change is At the goal hub we wait for a flag (variance in the not proportional) to the electric field E of the light. Normally, power of the flag) which lands at a specific moment(s) in time. this nonlinearity is observed only at very high light intensities In the event that there is no flag touching base right then and in the presence of extremely strong electric fields (eg by there, it implies that we have no answer for our concern. in the presence of Van De Graff Generator). Generally the issue has an answer. Vacillations can be perused But with the discovery of second harmonic generation with a and an oscilloscope. by Peter Franken in 1961, it is shown that second harmonic The main issue that is tackled along these lines was the generation can be created in many material such as gallium Hamiltonian way problem. Later, different issues have been selenide, Lithium iodate monopotassium phosphate and so handled using this method. implication of Second Harmonic Generation for Optical Computing. Advantages of Time Delay Optical computing: Second Harmonic Generation simply means the combination of 2 photons to generate a new photon of twice 1. They can be used in solving extremely complex the energy and twice the frequency of the progenitor photons. mathematical problems such as the Hamiltonian Path This in combination with optical resonations can be used in problem. transistors for performing logical operations. This has an 2. They can used to enhance cryptographic processes in added benefit of creating constructive interference which order to enhance security. enables the generation of light pulse with greater amplitude. 3. They can be made extremely fast and highly efficient This in theory could solve the problem of absorption of light compared to conventional computers. by the reflecting/refracting medium. An optical transistor can be made from 2 mirrors Disadvantages of Time Delay based Optical Computing: placed close to each other and the gap between them can be

Volume 7, Issue 3, 2018 38 http://dynamicpublisher.org/ International Journal of Scientific Research and Review ISSN NO: 2279-543X

3. The biggest disadvantage is that the circuits are 1. They can solve only one specific types of problems. extremely complex to design and build and the 2. They are extremely complex to create. complexity only increases with complexity of the 3. Although they are extremely fast compared to problem. traditional computers, they are very inefficient when it comes to power consumption and they also ISING MACHINE :- generate a lot more heat than regular silicon based computers (Due to the absorption of light by the One of the biggest problems that computer manufacturers are facing today is the increasing complexity of reflecting/refracting surfaces) the algorithms that is increasingly taking over the world around us. One of best examples of such problems is the WAVELENGTH BASED OPTICAL COMPUTING :- Travelling salesman problem. According to this problem a salesman is to conduct business in various cities across the In wavelength based optical computing, the county, can travel to each city only once. Although this seems computation done by the utilizing the wide range of light to a simple problem, it becomes more complicated as more wavelengths in order to solve NP Complete Problems such as and more cities are added to the network. In fact, it has SAT3. These types of problems as mentioned before become so complex, even super computers cannot solve this extremely difficult to solve by traditional computers. This problem quickly. But by using the Ising Machines it is unconventional form of optical computing is accomplished by hypothesized that this type of computer can be used to solve using various filters and optical resonators. The basic process such problems. behind this is: An Ising machine is nothing but a computer composed of , magnets and other reflecting/refracting medium. It 1. The various wavelength frequencies of light are uses an optical table in order to process data It consists of mapped to various value assignments or variables. magnets, lasers, mirrors and prisms which can be used to solve 2. The selected frequencies of light are obtained by extremely difficult calculations. passing the light through a Optical Filter Modules (OFM) that filters out unwanted frequencies of light OPTICAL FOURIER CO-PROCESSORS :- for obtaining a desired Boolean expression This form of optical computing paradigm uses the 3. Using the Boolean expressions and the OFMs to principle of Fourier in order to solve extremely difficult problems such as rigorous scientific experiments. These solve the given SAT3 problem. computers work by the describing a discrete Fourier transform [DFT] in a input signal by using a spatial light modulator and The light is first focused into the system by using a the result is measured using CMOS or CCD image sensor. series of prisms and reflectors. The various frequencies are Such optical architectures can offer superior scaling of then mapped to certain values, such as binary values. The light computational complexity due to the inherently highly is then sent to the OFM in which non-assigned wavelengths interconnected nature of optical propagation, and have been are filtered and then a Boolean expression is added to it. Then used in the past to solve 2D heat equations the wavelength and generated Boolean expression is used to solve the given SAT3 problem by equating the wavelength to the given SAT3 formula. The resulting data is then compared CONCLUSIONS :- to the Boolean expression and the problem is satisfied only when the value of the resulting expression returns ‘1’ or the Firstly, it is must be noted that this is not an exhaustive sensing of the light pulse by the photo-detector. comparison of the optical computing paradigms. There are For example, consider a problem which has multiple some that are not covered by this review as they are mostly variables. Let’s take the time required for a light beam to pass theoretical or can computing capabilities are less than the ones through a CNF designed circuit is time T. The shown method mentioned here. Although the research into fully optical takes time T to process each instance of the given problem. It computers is still in their nascent stages, many strides have must be noted that the no of prisms and optical dividers is also been taken in this field. Companies like Hewlett-Packard have equal to T. This implies that the complexity of the circuit is released a processor that computes using light signals. dependent on the time required to solve the given problem. Although we will be seeing a shift towards a combination of electronic and optical based devices, Optical computers can Advantages and Disadvantages of the Technology offer lot more than traditional Silicon based computers. Although many of\the top Computer Manufacturer and 1. They can solve extremely complicated mathematical Consumer Technology firms have shown an interest in development of Quantum Computers and Neural Networks, it calculations is pertinent to understand that Optical and other 2. They can be made highly efficient Unconventional Computing Technologies such as DNA or Wetware Computing and Chemical Computing can also be

Volume 7, Issue 3, 2018 39 http://dynamicpublisher.org/ International Journal of Scientific Research and Review ISSN NO: 2279-543X used in parallel to traditional computing. Although many [3] What Classical Optics can do for the Digital Optical Computers, Adolf challenges remain in the development of a fully Optical W Lohmann, Applied Optics 25, 1986, pp. 1543-1549. Computer, it is also important to understand that developments [4] Digital Optical Computing, K.H Brennes, Applied Physics B 45(2), 1988, pp. 111-120. and the various paradigms in Optical Computing can make it a [5] Optical Computing: A Survey for Computer Scientists, Dror G Fetelson, reality. It is important to understand that although there many MIT Preas, 1988. other paradigms that are not covered by this review, the ones [6] Three Dimensional Nanostructures as highly efficient generators of mentioned here have the strongest potential to be used in fully Second Harmonic Light, Yu Zhang, Nathaniel K Grady, Ciceron Ayala- Optical Computer. Of the ones mentioned above Time Delay Orzoco, Naomi J Halas, Nano Letters 11(12), 2011, pp. 5519-5523. based Computing and Wavelength based Computing are the most eligible for the task. We hope that the findings of this [7] Role of Optic in Computing, Rodney S Tucker, Nature 4, paper will be used in the development of a future Optical 2011, pp. 405. [8] A Light based Device for solving Hamitonian Path Problem, Mihal Computer. Olteans, Unconventional Computation, 2006, pp.217-227. [9] An Optical Wavelength based Solution to the 3-SAT problem, Aama Gobaei, Aaeed Falili, Optical Super Computing, 2009, pp.77-85. [10] An Optical Fourier Transform Co-Processor with Direct Phase REFERENCES Determination, AJ Mcfaden, SD Gordon, TD Wilkinson, Scientific Reports,2017. [11] A coherent Ising Machine for 2000 mode Optimisation [1] Digital Optical Computing, AD Aauchuck, TC Proceedings of IEEE, problem,Takahiro Inagaki, Yoshitaha Haribara, Koji Igarashi, 2016, 1984, pp.758-779. pp.354,603-606,6312. [2] Photorefracture Non Linear Optics and Optical Computing, Pochi Yeh, [12] Large Scale Ising optics network based on degenrate optical parametric Arthur E Cheon, John Hong, Paul Beckwrith, Tallis Chang, Monte oscillators, Takahiro Inagaki, Kensuske Inaba, Ryan Hamerby, Nature Khoshneusian, Land Mark Papers on Photofracture and Non Linear Photonics 10, 2016, pp.415-419 Optics, 1995, pp. 525-540.

Volume 7, Issue 3, 2018 40 http://dynamicpublisher.org/