A new photonic chip developed by researchers from MIT and Nokia promises to transform artificial intelligence computations by utilizing light instead of electricity.
This fully integrated photonic processor performs essential calculations for deep neural networks optically, achieving over 92% accuracy in under half a nanosecond—matching traditional electronic hardware.
Manufactured using standard foundry processes, the chip is scalable and can be integrated with existing electronic systems. It features a novel component called NOFU (Nonlinear Optical Function Unit), enabling both linear and nonlinear operations directly on the chip.
The team aims to integrate this technology with real-world electronics, such as cameras and telecommunications systems, and to develop algorithms that leverage optical advantages for faster and more energy-efficient training of AI systems.
With capabilities to deliver rapid responses in nanoseconds, this innovation opens new avenues for advanced applications in AI.