Researchers Develop Optical Chip Capable of Handling Negative Values for AI Computations

Researchers have developed a fully real-valued optical chip capable of processing light to perform computations, including handling negative values, for generative models. This advancement addresses the growing demand for computational power and energy efficiency as artificial intelligence (AI) algorithms become increasingly complex. The innovation leverages Optical Neural Networks (ONNs), which utilize light instead of electricity to process information, offering potential advantages in speed and energy consumption.

The newly designed optical chip introduces a method that allows light-based systems to handle negative values, a critical component for many AI tasks, particularly in generative modeling. Traditional ONNs have faced challenges in representing negative numbers due to the nature of light waves. By overcoming this limitation, the new chip expands the capabilities of ONNs, enabling them to perform more sophisticated computations required by advanced AI algorithms. Researchers highlight this development as a step toward more efficient and powerful hardware solutions tailored for the evolving demands of AI technologies.

Newsflash | Powered by GeneOnline AI

Source: GO-AI-ne1

For any suggestion and feedback, please contact us.

Date: February 5, 2026

Author

GOAI