Nature Publishes Xizhi Technology's Optoelectronic Hybrid Computing Achievement

Nature Publishes Lightelligence's Optoelectronic Hybrid Computing Achievement

Original Link: https://www.nature.com/articles/s41586-025-08786-6

On April 9th, London time, the world's top academic journal Nature published Lightelligence's optoelectronic hybrid computing achievement: "An integrated large-scale photonic accelerator with ultralow latency." This marks Lightelligenc's return to a top-tier global academic journal since Dr. Shen Yichen, founder of Lightelligenc, published the cover article "Deep learning with coherent nanophotonic circuits" in Nature Photonics eight years ago. This also represents another significant achievement for Lightelligencin the field of optoelectronic hybrid computing, following the release of its latest optoelectronic hybrid computing card, Xizhi Tian Shu (click to read), on March 25th.

This publication in Nature is another affirmation from the industry of Lightelligenc's successful development of the highly integrated photonic computing processor PACE (Photonic Arithmetic Computing Engine). Dr. Shen Yichen stated, "PACE is a product we released in 2021. We chose to publicly release its product details in a top journal like Nature to open up our technology roadmap, allowing more people to participate in this industry and accelerating the industrialization of the optoelectronic hybrid ecosystem."

PACE Optoelectronic Hybrid System

In recent years, with the development of disciplines such as silicon photonics, nano-optics, and materials science, the popularity of photonic computing has been continuously increasing globally. However, most research results are still in the laboratory stage. The primary reason why Lightelligence's productization achievement was accepted by Nature is that it demonstrated a large-scale optoelectronic integrated computing card manufactured on a commercial production line and provided all the measured data, confirming its significant advantage in computational latency.

The reviewers highly praisedLightelligence's team's efforts in the engineering of photonic computing: "In the field of photonic computing, it is common to make optimistic inferences about the performance of large-scale systems through small-scale demonstrations. However, the data in this paper all come from the measured performance of the entire PACE computing system. The authors have engineered an ultra-large-scale photonic matrix computing system, which is a remarkable achievement."

PACE is based on the fundamental principle of extremely low latency in optical matrix-vector multiplication. It achieves low latency through repeated matrix multiplication and cleverly utilizes tight loops composed of controlled noise, thereby generating high-quality solutions to combinatorial optimization problems such as the Ising problem and the Max-cut/Min-cut problem.

Furthermore, in the paper, Lightelligence also publicly disclosed its specific architecture for optoelectronic hybrid computing for the first time. During years of productization, Lightelligence's R&D team discovered that photonic matrix computing based on an incoherent architecture is more conducive to commercialization due to its advantages in precision control, matrix adjustment flexibility, and noise resistance. Wang Long, COO of Lightelligence, stated, "PACE is one of Lightelligence's three core technologies and a representative achievement of photonic matrix computing (oMAC). Since the advent of PACE, the Lightelligence team has continued to focus on improving the flexibility of optoelectronic hybrid systems and the efficient collaboration of optoelectronic chips, constantly exploring broader application scenarios."

PACE System Architecture Diagram

In 2017, Dr. Yichen Shen published a cover paper in Nature Photonics titled “Deep learning with coherent nanophotonic circuits,” proposing for the first time a method for deep learning computation using coherent nanophotonic circuits. This work is widely regarded as a pioneering advance in integrated photonics. That same year, Dr. Shen founded lightelligence, dedicated to productizing and commercializing photonic computing.

In 2019, lightelligence unveiled its first photonic computing prototype board, which successfully ran a convolutional neural network model included with Google TensorFlow to process the MNIST dataset, validating the feasibility of photonic computing. Since then, the lightelligence team has continuously explored scaling matrix sizes and improving photonic chip integration, while steadily strengthening engineering capabilities in electronic chip design, packaging, software, and systems.

The launch of PACE in 2021 marked a major leap for optoelectronic hybrid computing—from proof of concept to productization. In March 2025, the lightelligence Tianshu Compute Card was introduced (click to read). It quadrupled the photonic matrix size to 128×128 and addressed several key engineering challenges, including the first successful implementation of 3D TSV advanced packaging on an optoelectronic hybrid chip, as well as the miniaturization and onboard integration of the light source. This made Tianshu a standard full-height, full-length PCIe card that can be plugged directly into existing server hardware. More importantly, Tianshu was applied to complex commercial models for the first time and demonstrated latency advantages over commercial GPUs in specific algorithms, signaling that the commercialization of optoelectronic hybrid computing is on the horizon.

Dr. Huaiyu Meng, Co-founder and CTO of lightelligence, said: “We have always believed that optoelectronic hybrid computing will soon become a new mainstream computing paradigm, and we have consistently made productization and commercialization our development goals. We hope more application partners, researchers, developers, and collaborators will join the optoelectronic hybrid computing ecosystem to explore more application scenarios and accelerate the market adoption of optoelectronic hybrid products. lightelligence will also continue to share its latest R&D achievements with both industry and academia.”