A professor from Shanghai Jiao Tong University starts a business, using AI to develop new materials, and is invested by a fund affiliated with Shanghai Jiao Tong University | First exclusive report by 36Kr

Author: Ou Xue

Editor: Yuan Silai

Yingke learned that Suoge Zhisuan, a startup in the field of AI for Materials (AI-enabled new material R & D), announced the completion of a seed round of financing exceeding 10 million yuan. This round of financing was led by Qigao Capital, with Jiaoda Hanyuan Asset, Zizhu Xiaomiao, and Zizhu KeTou participating as follow - on investors. The funds will be mainly used for the continuous R & D of the original AI computing engine, the expansion of the R & D team, and the construction of computing power infrastructure.

Suoge Zhisuan was founded in September 2025. The founding team is from Shanghai Jiao Tong University, led by Xu Zhenli, a specially - appointed professor at the School of Mathematics and the director of the Research Center for New Materials in Artificial Intelligence. Other core members of the team also have interdisciplinary backgrounds in mathematics, artificial intelligence, high - performance computing, and materials science.

Yingke learned that in the field of new material R & D, traditional computational simulation methods often face the "impossible triangle" of "high precision - high efficiency - low cost". Especially, the long - range interaction in material simulation has always been a bottleneck restricting computational accuracy.

Suoge Zhisuan is committed to using algorithms to solve the industry pain points of long R & D cycles and high costs in new material R & D. One of its core technologies is the new neural network descriptor it proposed, SOG - Net. This algorithm decomposes the total potential energy into short - range and long - range terms for independent modeling and efficient coupling, introduces a trainable Gaussian sum function, and adaptively fits long - range interactions such as Coulomb and dispersion, significantly reducing the prediction error of the energy and force of complex systems.

Schematic diagram of SOG - Net (Source: the enterprise)

Meanwhile, based on the original algorithm, the team developed the random batch molecular dynamics simulation software RBMD and further launched a dedicated simulator integrating software and hardware, NanoTitan. This device can complete simulations of tens of millions of atoms on a single GPU card, with a computing speed dozens of times faster than mainstream software, shortening the R & D cycle of new materials from several years to several months.

NanoTitan all - in - one machine (Source: the enterprise)

Currently, the RBMD algorithm has been connected to the National Supercomputing Internet Platform, and the NanoTitan all - in - one machine has been put on the market, serving many universities, research institutes, and R & D institutions.

Although it has been established for a relatively short time, Suoge Zhisuan has initially formed a business model with both to C and to B operations. The to C end is mainly based on the NanoTitan all - in - one machine, targeting computing laboratories in universities and research institutes; the to B end focuses on the new material R & D needs of large enterprises, directly connected by the core team.

In addition, Suoge Zhisuan has formed a clear layout in key areas of new materials. In the direction of rare - earth permanent - magnet materials, the company will jointly build an AI R & D and application system for rare - earth permanent - magnet materials with the listed company Tianhe Magnetic Materials, forming a full - process closed - loop of "material design - experimental verification - engineering application".

In the field of lithium - battery materials, the team cooperated with the relevant team of the Future Energy Research Institute of CATL and proposed the R2D multi - field coupling model, breaking through the electrochemical modeling framework based on the homogeneous assumption in the past three decades. This model can accurately predict the failure mechanism of battery materials, providing theoretical support for the R & D of the next - generation high - safety and high - energy - density batteries.

In addition, Suoge Zhisuan has also established cooperation with Huawei and carried out joint R & D in the fields of high - performance computing and operator learning for domestic chips.

Application scenarios of Suoge Zhisuan's technology (Source: the enterprise)

After this round of financing, Suoge Zhisuan will continuously expand its core R & D team and gather global interdisciplinary talents. On the one hand, it will accelerate the engineering implementation of core algorithms in multiple scenarios (such as rare - earth permanent magnets, new materials, semiconductors, etc.), strengthen the technical barrier driven by "supercomputing + AI", and build an integrated platform for new material R & D; on the other hand, through resource aggregation, it will promote the company's transformation from "scientific research breakthrough" to "market development", laying a solid foundation for subsequent large - scale development.

Views of investors:

Qigao Capital said: "We firmly believe in the development potential of the AI for Materials track. The scale of China's new material industry is huge, and the AI penetration rate is still in its early stage. The upgrade of the industry's technological paradigm is giving rise to significant opportunities. The original algorithm breakthrough of Suoge Zhisuan's team in long - range interface simulation and the domestic high - performance molecular simulation all - in - one machine accurately solve the core pain points of 'precision - efficiency' in the industry, and have outstanding competitiveness in both scientific research and industrial sectors. The company's development path of algorithm innovation and software - hardware integration has great advantages and is expected to become a leader in the field of AI - enabled new material R & D, releasing important value in key tracks such as batteries, rare earths, and semiconductors. Qigao Capital will continue to provide support to help the company accelerate the industrialization of technology and market expansion."

Jiaoda Future Industry Investment Fund said: Suoge Zhisuan has achieved a breakthrough in the underlying algorithm in the field of AI + materials science with the top - notch interdisciplinary team led by Professor Xu Zhenli. Its core SOG - Net and random batch acceleration algorithm accurately solve the problem of long - range simulation of material interfaces and are efficiently implemented with the domestic software - hardware all - in - one machine, with a significant technological moat.

We are optimistic about the trillion - scale track of AI - driven material R & D. The company uses a differentiated path of "precise force field + dedicated hardware" to enter key fields such as solid - state batteries and rare - earth permanent magnets, with a clear commercialization prospect. The team combines in - depth scientific research with an industrial perspective and is a rare benchmark for hard technology.

We look forward to Suoge Zhisuan continuing to lead the new paradigm of "intelligent design", transforming academic excellence into industrial value, and becoming the core engine of China's new material innovation.

Zizhu Xiaomiao said: Suoge Zhisuan is a typical project for the transformation of scientific and technological achievements from Shanghai Jiao Tong University. Professor Xu Zhenli has been deeply involved in computational mathematics for more than 20 years. With the original "random batch" series of algorithms developed in the laboratory, he solves the large - scale bottleneck of molecular dynamics simulation at the underlying level. This kind of bottom - up innovation from 0 to 1 is exactly the value most cherished in seed - stage investment.

We appreciate the academic confidence of the team and look forward to their courage to cross the "valley of death": from the algorithms in papers to the "NanoTitan" all - in - one machine in customers' hands, from the laboratory of Shanghai Jiao Tong University to real industrial scenarios. AI + material design is not just a story on the wave of the moment but requires long - term technological breakthroughs. We look forward to Suoge Zhisuan becoming an important force in the AI + material design track.

Zizhu KeTou said: As a companion for Suoge Zhisuan's development from the concept verification stage at Shanghai Jiao Tong University to industrialization, we firmly believe that with the requirements put forward in China's "15th Five - Year Plan" to "fully implement the 'Artificial Intelligence +' initiative, lead the transformation of scientific research paradigms with artificial intelligence, strengthen the combination of artificial intelligence and industrial development, and empower all industries in all aspects", combined with the considerable domestic market pattern of new materials, the Suoge team has achieved breakthroughs in intelligent algorithms and accumulated industry experience. They have cooperated and conducted joint research with domestic leading enterprises such as CATL and Huawei in the fields of new energy and rare earths, demonstrating the technical capabilities and industrial value of a first - class university's scientific and technological team. In the future, with AI + as the core, the enterprise will build a system of "material R & D, pilot - scale production and maturation, and material intelligent manufacturing", which will further help China break through the R & D bottleneck of new materials and is expected to promote and lead the optimization and upgrading of domestic advanced basic materials, key strategic materials, and frontier new materials. Zizhu KeTou will continue to carry out industrial collaboration, focus on key breakthroughs, and strive to transform the concept verification project at Shanghai Jiao Tong University's Zizhu campus into an industry "unicorn" enterprise.