Jump to content


Graphene-based All-Optical Technology Platform for Secure Internet of Things

The need for a next-generation computing platform becomes clear from IoT and 5G/6G and their high performance and low power requirements. Now, graphene and 2D materials (2DM) offer the unique ability to enable highly confined nonlinear interactions of light at low powers and at extremely low response times in the femtosecond range. However, it must be integrated with CMOS low-loss silicon nitride (SiN) platform that facilitates the possibility to create circuits for fast, low power, high bandwidth, general purpose computing and memory completely in the optical domain.

As the most important challenge comes from the maturity of the graphene processes with standard CMOS environments, the main goal of GATEPOST is to fabricate and demonstrate a radically new graphene-based all-optical data processing platform, integrated and tested in a real CMOS pilot line.

As a user case, we focus on a network security device for distributed denial of service (DDoS) detection and network packet inspection. Even though on average 170 cyber-attacks are performed per IoT device per day, there is still a huge lack of security due to the added power, latency, operating costs and bandwidth limitations involved. This is unacceptable, especially considering that the cybercrimes topped an estimated $318 billion in 2021 alone. With our graphene-based computing platform, we will show how low-power, low-latency and high bandwidth network security is ready for the IoT and 5G/6G future.

The full system showcases the unique expertise of each consortium member in all-optical digital logic, neuromorphic computing, memory and ultra-fast clock generation. These components are realized in the 2D-Experimental Pilot Line at IHP, allowing for scalable fabrication and strengthening the EU’s supply-chain in high-performance computing. In the future, the developed platform can be deployed for applications in AI, autonomous driving and more, paving the way for computing beyond von Neumann and Moore’s Law.