Tsinghua University Introduces Distributed Broadband Light Computing Architecture, Unveils Tai Chi Chip

On April 12, according to reports from the media, Tsinghua University pioneered a distributed broadband light computing architecture and developed the Tai Chi chip, a large-scale interference-diffraction heterogeneous integrated chip, achieving a universal intelligent computation with 160 TOPS/W efficiency.

This research breakthrough discards the traditional paradigm of electronic depth computation, embarking on a journey through the realm of photons to explore new inspirations, architectures, and pathways for high-performance computing.

With the flourishing development of artificial intelligence, intelligent light computing, as a novel computing modality, demonstrates performance and potential far beyond silicon-based electronic computation in the post-Moore’s Law era.

However, the major challenge lies in the fact that the computational advantages of light are trapped within unsuitable electronic architectures, limiting the scale of computation and hindering the support for complex large-model intelligent computations that urgently require high computational power and efficiency.

The computational efficiency of the Tai Chi light chip directly surpasses existing intelligent chips by 2-3 orders of magnitude, providing computational support for tasks such as high-speed intelligent analysis of billion-pixel large scenes, training and inference of billion-parameter large models, and milliwatt-level low-power autonomous intelligent unmanned systems.

[End of Article] For reprinting, please indicate the source: Quick Technology