Abstract
The continued performance scaling of AI gigafactories requires the development of energy-efficient devices to meet the rapidly growing global demand for AI services. Emerging materials offer promising opportunities to reduce energy consumption in such systems. In this work, we propose a novel electro-optic microring modulator that exploits a graphene (Gr) and transition-metal dichalcogenide (TMD) interface for phase modulation of data-bit signals. The interface is configured as a capacitor composed of a top graphene layer and a bottom WSe2 layer, separated by a dielectric Al2O3 film. This multilayer stack is integrated onto a silicon (Si) waveguide such that the microring is partially covered, with coverage ratios varying from 10% to 100%. The proposed device design and its key figures of merit, including energy efficiency, are analyzed. Simulation results indicate that the microring modulator achieves low energy consumption and high-speed operation while maintaining a compact footprint. In particular, the device operates at 39.1 GHz with an energy consumption of 8.3 fJ/bit under 25% Gr–TMD coverage, whilst occupying an area of only 20 μm2. Furthermore, a modulation efficiency of VπL= 0.082 V·cm and an insertion loss of 6.8 dB are obtained for the 25% coverage. The proposed Gr–TMD-based microring modulator demonstrates significant potential for high-speed, energy-efficient data modulation, contributing to the development of more sustainable AI gigafactories.

This publication has 38 references indexed in Scilit: