Optics Communications 483 (2021) 126668 Contents lists available at ScienceDirect Optics Communications journal homepage: www.elsevier.com/locate/optcom A review on all-optical logic adder: Heading towards next-generation processor Kamanashis Goswami, Haraprasad Mondal ∗ , Mrinal Sen ∗ Department of Electronics Engineering, Indian Institute of Technology (Indian School of Mines), Dhanbad, Jharkhand, 826004, India ARTICLE INFO Keywords: Photonic Crystal All-optical effect Ring resonator Semiconductor Optical Amplifier Plasmonics Mach–Zehnder interferometer ABSTRACT In this paper few literatures on all-optical adders, reported approximately in last fifteen years, have been reviewed. Being the key element of arithmetic and logic unit (ALU), adder has attracted stirring attention of the researchers. The reviewed articles are primarily categorized into three broad sections based on the platforms utilized in designing all-optical adder namely semiconductor optical amplifier (SOA), plasmonics and photonic crystal (PhC). SOA based designs offer high contrast ratio between the output logic levels, however; at the cost of high response time and operating power. Plasmonic based designs provide miniaturized physical footprint of the devices owing to the tight optical confinement. Nevertheless, the process of fabrication involved with the plasmonic based designs along with the loss incurred at the metal are among the limiting features. In contrast, PhC based designs are much more advantageous owing to the notable characteristics of light propagation. The operating power and bandwidth of operation are also higher as compared to the other designs. Photonic crystal- based designs are further categorized into linear and nonlinear domains based on their operating principle. The design structures and performances have been summarized and a comparison is made. Some insights have been discussed leading to design of more efficient PhC based all-optical adders for next generation ultra-first optical processors. 1. Introduction Emerging need of high-speed data processing has made applications of all-optical signal processing circuits and logic devices inevitable in future generation computing platforms. Leading technology providers, including INTEL, IBM, LUXTERA etc., have already started to adopt the technologies offered by Photonic Integrated Circuits (PICs) [1– 3] to keep up with the Moore’s law. Global market of PICs, even being severely hit by the pandemic COVID-19, is expected to reach a compound annual growth rate (CAGR) over 26 percent by the middle of the upcoming decade [4]. The primary reason behind this expectation to PIC is its unprecedented combination of merits like — high (rather super!) speed and low power consumption. The secondary reason (how- ever, foremost in economic viewpoint) is the ability of silicon to be used as a promising optical material whose foundry is well-established owing to the matured CMOS technologies. Research on PICs has got advanced through its several tributaries. Although each of them has their own specialism; they, based on their principle of operation, can broadly be classified as linear or nonlinear optical devices. Both of these clusters have their own limitations as well as merits. However; researches on PICs, irrespective of their cate- gory, have never got stuck in fulfilling the expectations of demanding technologies by providing with myriads of optical devices ranging ∗ Corresponding authors. E-mail addresses: kamanashis.goswami@gmail.com (K. Goswami), mandal.haraprasad@gmail.com (H. Mondal), mrinal.sen.ahm@gmail.com (M. Sen). from sensors [5,6], multiplexers/de-multiplexers [7–10], switches [11], logic gates [12–16], decoder [17–22], polarizer [23], coupler [24] to processors [25]. Among these, a vast community of researchers and technology professionals are putting their eagle eye on all-optical logic gates, processors and its subsections; on account of their potential use in next generation high speed computations. Therefore, as obvious, numerous promising solutions have been proposed in these contexts during the last few decades. In spite of that, fully functional all-optical processor is yet far from the reality. Nevertheless, some of its important building blocks, like - basic Boolean logic gates (including NOT [26,27], AND [28–30], OR [31,32], NAND [33,34], NOR [35], XOR [36] etc.) and ADDER, have already paved through the aisles taking them out of their infancy. In these, the ADDER is an unavoidable block towards realizing the Arithmetic and Logic Unit (ALU) of a processor. An all- optical ADDER (AOA), that can efficiently restore logic levels at its output as that at its input; offers cascadability; is less power hungry; incurs minimal criticality in fabrication; and occupies a small footprint, is expected to fulfil the dream of realizing a fully functional optical processor. In this context, a neophyte of the domain always needs to know how close we are towards the reality of an AOA. Attempting to light on this line of thought, this work endeavours to put down the important works in this topic under a single umbrella in form of a review. https://doi.org/10.1016/j.optcom.2020.126668 Received 10 October 2020; Received in revised form 26 November 2020; Accepted 30 November 2020 Available online 3 December 2020 0030-4018/© 2020 Elsevier B.V. All rights reserved.