MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202641063872 A) filed by Boray Sudhindra Deepaksubramanyan on May 20, 2026, for Nemsformance - Photonics On-Chip.
Inventor includes B. S. Deepaksubramanyan.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: 1. NEMSFormance - Photonics on-chip : A photonic interconnect medium for computing systems called NEMSFormance is presented [0][1]. With a focus to implement high-performance computing systems, this new quantum dots (artificial atom, scalable) and nanocantilever based manycore on-chip point to point interconnect medium (can supports electrical as well as photonic signalling) is explained here with the premise of photons are faster than electrons. This goal of NEMSFormance is to facilitate faster intra- and inter core data transfers, as well as 3D-stacked computing systems. Nanoelectromechanical (NEMS) slab (with actuator when possible) is proposed to be used to deflect the emitted photon(s) to various compute cores depending on the selected architecture. This NEMS slab would require an Aluminium coating that to the best of this author’s knowledge and understanding is able to reflect the required range of wavelength (energy). Applications include 3D-Processors [2] (possibly zetta scale computing), quantum Science and quantum information processing. 2.MemTower – A Photonic Memory Unit for 3D-Photonics On-Chip : A NEMS cantilever based photonic memory unit with quantum dots called MemTower is presented. Nanocantilever of around 10nm length (possible?) built with silicon and aluminium forms the primary basis for design of the proposed photonic memory unit. FeSi compounds (ferromagnetic) based Quantum dot is used in this memory unit to store data. Working of the proposed memory unit is based on theory of deflection of photon by means of reflection. Nanocantilever beam with a fixed end is placed above the quantum dot that stores data. Spontaneous emission from the quantum dot is deflected by reflecting the emitted photon with Aluminium (or silver) coating based Si nanocantilever which can be bent with Nickel or PZT based material deposition to ensure appropriate direction of photon (that represents data) travel to a computing unit in a computing system. As photons are faster than electrons, Latency is reduced which leads to faster computing, the purpose and goal of this work. With this memory unit, 3D-Integrated Circuits implementation that can be both face-to-face (F2F) as well as face-to-back (F2B) is shown with memory- on-logic (MoL) as a reference implementation.
Disclaimer: Curated by HT Syndication.
