Imagine a computer chip that thinks like a human brain, using quick pulses instead of a constant stream of data. This project mixes light-speed signals with energy-saving electrical switches on a single chip. It's like combining a sprinter's speed with a marathon runner's endurance to process information faster and with much less power.
Current AI hardware is hitting a wall regarding energy efficiency and processing speed. Traditional chips cannot handle the demands of real-time, high-speed data processing without consuming excessive power.
An integrated circuit combining laser optical neurons and electrical spiking diodes. The team has optimized fabrication processes for active laser materials and grown wafers for electrical devices.
If you are a robotics manufacturer dealing with lag in high-speed control systems — this project developed a photonic-electronic integrated circuit that enables ultra-fast and efficient processing. This allows machines to react in real-time with lower energy costs.
If you are an edge computing provider dealing with high battery drain in AI sensors — this project developed a chip that reduces energy consumption per spike. This extends the operational life of remote sensors processing data on-site.
If you are a research equipment firm dealing with the inability to simulate complex brain activity in real-time — this project developed neuromorphic hardware for computational neuroscience. This provides a hardware platform that mimics biological spiking neural networks.
Based on available project data, there is no specific pricing or cost information provided as the project is in the research and development phase.
The project focuses on a common technology platform for chip integration, but based on available project data, industrial scaling metrics are not yet defined.
Based on available project data, specific licensing terms are not listed; however, the project involves a consortium of 6 partners including a university coordinator.
The project is developing new learning strategies and algorithms specifically designed to work with the hardware parameters of the electrical and photonic devices.
The project period is from 2024-03-01 to 2028-02-29.
The consortium is heavily research-oriented, consisting of 4 universities and 1 research organization, with only 1 industry partner (17% industry ratio). This suggests the technology is in an early stage of development, focusing on fundamental breakthroughs in photonic-electronic integration rather than immediate commercial rollout.
Contact the Technical University of Eindhoven (TUE) regarding the SPIKEPro project.
Contact us to track the transition of SPIKEPro from lab to fab.
