"X-Day" Xili Lake Roadshow Club × "Venture China" AI for Science Special Session Visits Tsinghua University
From Haidian, Beijing to Nanshan, Shenzhen, from Tsinghua University laboratories to Shenzhen pilot production lines, AI for Science (AI4S) is becoming a key variable of new quality productive forces. This event gathered more than 200 representatives from government authorities, university research teams, leading venture capital institutions, industrial capital and industrial service providers, to open up the full-link channel for source innovation in universities, leading venture capital and industrial implementation, and seize the development dividend brought by the AI industrial transformation. The content is generated by AI and for reference only.
From Haidian, Beijing to Nanshan, Shenzhen, from Tsinghua University laboratories to Shenzhen pilot production lines, AI for Science (AI4S) is becoming a key variable of new quality productive forces. On the afternoon of October 8, 2026, the "X-Day" Xilihu Roadshow Society joined hands with the "Venture Capital China" Sci-Tech Roadshow Hall to bring the AI for Science special session to Tsinghua University Campus. This event gathered more than 200 representatives from government authorities, university research teams, leading venture capital institutions, industrial capital and industrial service providers, to open up the full-link channel for source innovation in universities, leading venture capital and industrial implementation, and seize the development dividend brought by the AI industrial transformation.
Industry Leaders Gather at Tsinghua University to Explore Opportunities in the AI4S Era
At the event site, Liu Minqiang, Director General of the Investment Promotion Agency of the Ministry of Commerce, Li Zhina, Vice District Mayor of Nanshan District of Shenzhen, Peng Gang, Member of the Standing Committee of the Party Committee and Vice President of Tsinghua University, Hu Chuxiong, General Coordinator of Innovation and Entrepreneurship and Deputy Director of the Academic Affairs Office of Tsinghua University, Ni Zhendong, Founder and Chairman of Zero2IPO Group, Chairman and CEO of Zero2IPO Holdings, Li Zhu, Founding Partner of Inno Fund, Yang Jianxin, Director of the Fundamental Industry Training Center of Tsinghua University, Zhang Qi, Head of the Mass Entrepreneurship and Innovation Teaching Department of the Fundamental Industry Training Center of Tsinghua University, Hua Hua, Partner and Chairman of Huakong Fund, Ji Wei, Founding Managing Partner of China Renaissance Capital, Wang Xuehui, Managing Partner of Shuimu Alumni Fund, Liu Bo, General Manager and Managing Partner of TusStar Venture Capital, Yan Yi, Partner of Shuimu Venture Capital, Gao Yuzhu, Partner of Qingyuan Capital, Xu Changzhong, Founder of the Listed Company Entrepreneur Exchange Center and Chairman of Huiyin Deqing Fund, as well as representatives from government, academia and investment circles including GigaDevice and Zhiyi Investment gathered together. At present, AI for Science is regarded as the fifth major scientific research paradigm following experiment, theory, computation and data. With the introduction of relevant supporting policies in Beijing and the advancement of the construction of the Zhongguancun (Haidian) Scientific Intelligence Innovation Cluster, AI4S has risen from a hot spot of academic research to a national strategic direction and important infrastructure for industrial development, with its applications in new materials, biomedicine, quantum information and other fields continuing to heat up. How to solve common problems in the industry and find the correct path for commercial implementation has become the core topic discussed by the guests on site.
In-depth Dialogue: Disassembling AI4S Technology Iteration and Commercialization Path
Hosted by Liu Bo, General Manager and Managing Partner of TusStar Venture Capital, front-line investors including Hua Hua, Partner and Chairman of Huakong Fund, Ji Wei, Founding Managing Partner of China Renaissance Capital, Li Zhu, Founding Partner of Inno Fund, and Wang Xuehui, Managing Partner of Shuimu Alumni Fund, carried out in-depth dialogues on the AI for Science technology iteration trend and commercialization path, facing the track opportunities and practical pain points directly.
Hua Hua, Partner and Chairman of Huakong Fund, pointed out that the current industry faces multiple challenges: first, how to improve the prediction accuracy of vertical domain models based on large models; second, how to form a complete closed loop of dry and wet laboratories; third, the reliability of data accumulation and the degree of "AI ready" are far from sufficient, and how to cross this stage is an important challenge. On the whole, AI for Science is still in the early stage, and the industry in the future needs to achieve systematic breakthroughs in model technology, full closed loop of experimental conditions, and overall data construction. At the project investment level, he focuses on the track market space, technology maturity, and business model implementation path. At the same time, he emphasizes that in this round of hard technology entrepreneurship wave, the comprehensive ability and entrepreneurial spirit of the founding team are the core assessment indicators.
Ji Wei, Founding Managing Partner of China Renaissance Capital, compared the development differences between the two core tracks of materials and pharmaceuticals from the dimension of commercial implementation. She pointed out that the demand in the materials track is clear, but it is relatively more difficult to manifest the value of AI R&D to end products, with a single paying subject and many industrial constraints, making commercial implementation more difficult. The pharmaceutical industry has a mature industrial division of labor and asset transaction system. AI can empower molecular screening and increase the probability of successful drug development. Enterprises can realize monetization through phased transactions at different clinical stages, which is a track with prominent commercial value. She also pointed out that there is a significant "island effect" in industry data, high-quality trainable data is scarce, and a large amount of data is scattered in laboratories and industrial ends. Even large manufacturers face the problem of data acquisition, which leaves opportunities for start-ups; AI for Science models cannot simply rely on "bigger effort brings better result", and must deeply integrate disciplinary mechanisms. How to complete high-quality data collection, screening and training is a problem that all entrepreneurial teams need to overcome.
Li Zhu, Founding Partner of Inno Fund, believes that against the background of the overall contraction of the venture capital market, AI for Science is one of the few high-quality tracks that continue to grow upward. The core can be summarized as four keywords: intelligence, breaking the upper limit of large models; computing, covering supercomputing and quantum computing; data, vertical experimental data leaves more space for start-ups; closed loop, realizing the closed loop of dry and wet experiments, and combining embodied intelligence to realize technology self-iteration. Taking new materials as an example, he pointed out that AI technology has compressed the material R&D cycle from the traditional ten years to one year, helping the domestic industry to achieve overtaking on curves. In terms of team quality, he values forward-looking vision and pattern, including the spirit of sharing, the spirit of contract and an open learning attitude; teams with unique data and model capabilities have more core competitiveness to survive through industry cycles.
Wang Xuehui, Managing Partner of Shuimu Alumni Fund, pointed out based on practical early-stage investment experience that the commercialization of different sub-fields of AI for Science varies significantly. The mathematics and physics fields mostly reflect the verification of model capabilities, and it is difficult to achieve profitability in the short term; pharmaceuticals and materials have more implementation possibilities, and start-ups can choose to provide technical services, and develop products independently when conditions are mature. Although industry giants also deploy AI+ science, their iteration speed lags behind relatively, leaving a window period for start-up teams. On the whole, the number of enterprises in the track that can eventually grow into large listed companies is limited, and the rest of the projects can be merged and acquired at the right time or cooperate with large manufacturers to achieve exit.
At the end of the roundtable, Liu Bo, General Manager and Managing Partner of TusStar Venture Capital, concluded that AI for Science, as a new generation of productivity tool, is being explored and applied in cutting-edge technology fields such as pharmaceuticals, materials, industrial manufacturing, rsi, quantum, and nuclear energy. She pays more attention to the breakthrough innovation brought by the cross-integration field, such as the interpretability and popularization of traditional Chinese medicine. Interdisciplinary cross-integration is the core key for this track to achieve breakthrough innovation. In the future, universities need to continue to deepen interdisciplinary basic education and scientific research innovation, continuously cultivate compound talents, and help the industry grasp the new development opportunities of AI empowering scientific innovation.
Project Roadshow: The Technology-Industry Spectrum Emerges Initially
In addition to the collision of ideas, this special session also carried out project roadshows simultaneously, focusing on showing the real industrial spectrum of the AI4S industry. The 6 participating roadshow projects cover key directions such as scientific computing power, AI chips, material agents, autonomous laboratories, virtual cells, and virtual organoids.
Quantum Kun Technology: An innovative enterprise that reshapes the scientific R&D paradigm with quantum artificial intelligence, committed to integrating quantum computing (Quantum), artificial intelligence (AI) and high-performance computing (HPC), building a future-oriented scientific computing infrastructure and ecosystem, and promoting practical quantum computing.
Xinyi Technology: An innovative high-tech company with core technology and independent R&D capabilities, committed to becoming a full-scenario AI inference chip provider based on RRAM storage-computing integration technology. The company's main business is AI inference chips and storage products based on new resistive random access memory (RRAM), mainly including full-scenario AI chips covering cloud, edge and end-side inference, embedded RRAM IP and stand-alone memory. Its application fields cover consumer electronics, security chips, medical electronics, network communications, industry and automotive sectors.
He Guang Zhi Cheng: An advanced material innovation enterprise that took the lead in proposing the full closed-loop model of "material design and synthesis agent + AI decision-making intelligent laboratory + AI chemical data infrastructure". Focusing on the field of molecular design and synthesis, it takes the AI molecular generation algorithm model as the brain, the high-throughput autonomous laboratory of AI decision-making as the hands, and the high-quality, high-precision, standardized and verifiable AI-Ready data as the blood, forming a disruptive material R&D paradigm that can truly meet industrial needs and open up industrial paths, ushering in the era of AI creation in material R&D.
Deep Autonomy: Exploring a new paradigm of autonomous scientific discovery driven by artificial intelligence and real experimental feedback for AGI for Science.
Shuquan Gene: Focusing on the core track of AI pharmaceutical services, it builds an AI virtual cell for dynamic disease evolution based on RNA. Its core technology uses plasma RNA as the precise dynamic molecular input to fully capture the panorama of cellular molecular expression; combined with the self-developed interpretable and perturbable AIVC digital twin cell engine, it builds a full-process technical system of "molecular analysis - target screening - cell simulation - efficacy prediction". It can be directly applied to clinical scenarios to avoid idle technology iteration; further, through personalized dynamic treatment strategies and molecular generation engines, it realizes virtual experiments, opens up the integrated diagnosis and treatment link, reduces the clinical cost of drug R&D, and forms a commercial closed loop of clinical-pharmaceutical factory services through "clinical scientific research - drug trials - pharmaceutical factory cooperation". It uses the clinical sample data flywheel to drive model evolution, forms a dynamically dimension-reduced AI iteration path, and creates a benchmark where business drives model evolution.
Yu Sheng Zhi Yuan: Committed to building a "digital life infrastructure" for drug development and precision medicine. With the highly simulated virtual organoid world model as the core, the company closely couples real organoids, patient-derived sample data and small animal immune models, monitors the dynamic fluctuations of immune markers in rare trace samples through self-developed single-molecule digital immunoassay technology, continuously calibrates and trains the immune microenvironment in virtual organoids, and continuously approximates the real human response in efficacy and immune safety prediction, gradually moving from virtual organoids to generalizable individualized digital life, which can be used in key scenarios such as tumor, neurodegenerative disease, drug resistance mechanism research, immunotherapy design, and epidemic prevention and control, accelerating new drug R&D and precision medicine implementation.
Strong Alliance: "X-Day" × Venture Capital China × Tsinghua University
This event is jointly built by two high-quality sci-tech innovation service platforms. The Fundamental Industry Training Center of Tsinghua University and the Tsinghua University Student Maker Space Association, as co-organizers, participated in depth. The three parties collaboratively build an accurate, efficient and full-link connected hard technology achievement transformation and capital docking channel.
The "X-Day" Xilihu Roadshow Society is a national sci-tech financial service platform created by Nanshan District. Through the full-process service of "precise screening, professional tutoring, roadshow docking, and long-term tracking", the platform forms an efficient closed loop, realizes the precise docking of financial resources and technology enterprises, and continuously improves the efficiency of serving the real economy. Since its launch, 21 themed roadshows have been held, attracting more than 581 high-quality projects to sign up, helping 62 enterprises obtain equity investment of over 3.556 billion yuan.
The "Venture Capital China" Sci-Tech Roadshow Hall is jointly initiated by the Investment Promotion Agency of the Ministry of Commerce, Zero2IPO Holdings, *Caijing* Magazine, Listed Company Entrepreneur Exchange Center and other parties. It closely follows the national future industrial layout, links national team capital, listed industrial groups, and front-line market-oriented VCs, and builds a high-credibility, cross-regional hard technology financing channel.
Relying on the resources of the two platforms and Tsinghua University's source scientific research and interdisciplinary innovation capabilities, this AI for Science special session has realized the efficient connection of top university scientific research achievements, industrial implementation resources, and leading capital markets, fully activating the innovation vitality of the track, and consolidating the docking foundation for the industrialization and commercial implementation of cutting-edge scientific research achievements.
It is reported that the next "X-Day" Xilihu Roadshow Society Overseas Innovation Project (APEC) Special Session will be held in Nanshan, Shenzhen on November 5, 2026. Investment institutions and overseas sci-tech innovation projects are welcome to log in to the official event website https://events.pedaily.cn/customized/1219/ to sign up for participation.
