involving tens if not hundreds of billions of transistors. This tremendous scaling of semiconductor devices is still predicted to continue for at least one more decade. [2] However, many obstacles begin to arise including challenges associated with physical effects, such as energy dissipation and spatial interconnectivity [3,4] which are largely caused by the principal foundations of the technology relying on transistors and/ or logic gates. Indeed, transistors are elementary switching devices that control the flow of electrical currents under the effect of voltages applied to their gates. However, these gates, together with some dielectric under them, form parasitic capacitances that get charged/discharged during the dynamic switching involved in computations. Such operations take time and consume energy from the available sources that are usually limited (i.e., batteries), creating boundaries and constraints both in the speed and efficiency of the available computing systems. [3] In this work we liberate ourselves from relying fully on charge-based devices in performing switching operations and look for alternative ways of unleashing the potential of electromagnetic (EM) waves for high-speed computing. EM waves already contribute to communications in computer systems but all the required switching is done using MOS-FET or semiconductor-based circuits. [5] Hence one may ask, is there a way of bridging this gap between EM waves and computing systems? to answer this question, outstanding efforts have been developed in the field of photonics and plasmonics for optical quantum computing [6,7] as well as beam splitting, matrix inversion, resonant plasmonic flow networks, [8-11] and 1D lattices using solitons [12,13] (representing an interesting example of a collision-based computing approach). Artificial electromagnetic media (metamaterials and metasurfaces as their 2D version) have also demonstrated to be great candidates for improved and new applications by controlling the EM response of media both in space and time. [14-19] Recently, they have been proposed to solve mathematical operations such as integrals, convolutions, and derivatives [20-23] as analogue computing devices. At their current research stage, such structures are potentially high-cost to implement and have mainly been proposed to solve a fixed computational operation, which could limit their future development for general-purpose computing systems. In our computing approach, however, we are not focused on performing arithmetic operations on data (as these can be performed by means of the technology at hand). Our technique relies on achieving an arbitrary control of the propagation of square EM Communication and transfer of information from one block to another within a system is fundamental for high-speed and efficient computing. Herein, a simple approach for computing, without using conventional electrical charge/ discharge-based primitive operations, in which information is represented in electromagnetic energy steps travelling in sections of tran...