Shulou(Shulou.com)12/24 Report--

On December 7, the 2023 volcano engine AI for Science closed-door seminar with the theme of "AI Drive, Science launch" was held in Beijing. The meeting invited dozens of experts from Shengxin, education, and other industries to jointly look forward to the new direction of the integration of artificial intelligence and traditional science, and discussed the new mode of production and integration of biopharmaceutical, chip, materials, industrial manufacturing, education, and other industries under AI4S.

Sun Zhirong, professor of the School of Life of Tsinghua University, winner of the China Bioinformatics Lifetime Achievement Award and head of the core group of the Preparatory Committee of the Chinese Society of Bioinformatics, said in a speech at the conference that in recent years, Al for Science technological achievements have broken out centrally, AI has played a great potential and important value in the field of natural science, and the emergence of large models is not only a major breakthrough in the field of deep learning. It also brings new opportunities for bioinformatics.

Sun Zhirong, professor of the School of Life of Tsinghua University and winner of the China Bioinformatics Lifetime Achievement Award, head of the Core Group of the Preparatory Committee of the Chinese Bioinformatics Society

Zhang Zemin, academician of new subjects of the Chinese Academy of Sciences, national special professor, Changjiang scholar, and director of Peking University's Biological Frontier Innovation Center (BIOPIC), also pointed out that the intervention of artificial intelligence has brought unprecedented opportunities for bioinformatics research, not only improving our understanding of life sciences, but also bringing new possibilities for fields such as disease treatment and gene editing.

Zhang Zemin, new academician of the Chinese Academy of Sciences, national special professor, and director of the Biological Frontier Innovation Center (BIOPIC) of Peking University, a Changjiang scholar.

At the meeting, the volcano engine also released the volcano engine scientific research smart cloud solution for the first time. The volcano engine will provide scientific research institutions and practitioners with full-scene products and solutions from IaaS to PaaS and then to SaaS, helping scientific research accelerate to AI4S.

Full-scene products and solutions are launched, and Scientific Research Intelligence Cloud provides full-chain support for AI4S

Volcano engine Scientific Research Smart Cloud solution Architecture Diagram

With the theme of "New Scientific Research Paradigm driven by AI", Zhang Xin, vice president of volcano engine, deeply discussed with the participants how to make better use of AI technology to promote the innovation and development of scientific research and improve the quality and level of scientific research results. Zhang Xin said that at present, AI technology is developing rapidly, which can provide new perspectives and solutions for scientific research, and has great potential, so AI4S has become a cutting-edge trend of the integration of AI and traditional science.

At the same time, the volcano engine has released a scientific research smart cloud solution. The solution is oriented to the pan-AI scenarios such as bioinformatics, health, artificial intelligence, material science, molecular simulation and EDA, and builds the capability of AI4S full-scene products and solutions, including infrastructure layer, platform layer and scientific research application layer, to help experts in the field of basic science unlock new discoveries.

Zhang Xin, vice president of volcano engine

Zhang Xin details the capabilities provided by these three-tier architecture:

In the infrastructure layer, volcanic engines can not only provide computing, containers, storage, networks and other infrastructure for scientific research through the public cloud, but also integrate the existing facilities of scientific research institutions through volcanic engine hybrid cloud veStack to create a hybrid cloud infrastructure.

In the platform layer, researchers can build a large model of academic research through the volcanic engine large model service platform Volcano Ark and pre-training model, and complete scientific research knowledge retrieval, paper intensive reading, data analysis Agent, paper writing assistance and other work with the aid of AI, so as to help researchers carry out academic research more efficiently and accurately, discover knowledge and make scientific research decisions. In addition, with the more flexible, excellent and self-service computing services provided by the volcano engine cloud high-performance computing service VECTOR, researchers can efficiently manage and flexibly use cloud-based high-performance computing resources

In the scientific research application layer, in order to meet the needs of efficient management of computing resources, dataset resources and applications, the volcano engine aims to reduce the threshold of computing-intensive scientific research, and creates a biological big data operating system (Bio-OS), which integrates automatic workflow choreography and Workspace, which helps scientific research computing from limited to infinite, and helps data / workflow management from isolation to interconnection.

With the acceleration of AI4S, volcano engine provides full-link AI4S full-scene products and solutions for researchers, which can not only help researchers manage and complete scientific research tasks more efficiently, but also effectively reduce the cost of scientific research.

The practice of four scenarios can provide reference for AI4S scientific research.

Volcano engine has been closely watching the development of AI4S. In the past year, volcano engine has cooperated with many partners in the field of AI, such as Shengxin, education, and so on, and achieved remarkable results.

Li Yixue, a researcher at Guangzhou Laboratory, explained the cooperation between Volcano engine and Guangzhou Laboratory from the Shengxin scene. With the help of volcanic engine Bio-OS, Guangzhou Laboratory completed the construction of a large scientific device, which solved the challenges of weak expansion of computing resources, complex tools, poor portability of processes and algorithms, and difficulties in collaboration, sharing and inheritance. In addition, Bio-OS also supports the development of CoT2G-F deep collaborative attention model in Guangzhou Laboratory, which can accurately predict the fitness and evolution of viruses, showing new opportunities for learning theoretical biology and related concepts and specific applications.

Li Yixue, a researcher in Guangzhou laboratory

Wen Han, head of research and development of macromolecules in Deep potential Technology and a researcher at the Beijing Institute of Scientific Intelligence, said that based on the GPU accelerated scientific computing solution provided by the volcano engine, Deep potential Technology has completed the training of Uni-RNA 's new context-aware deep learning model, which pre-trains the largest RNA sequence data set on an unprecedented scale, allowing the community to release the power of AI at a new level. Significantly accelerate the pace of scientific research, and then promote breakthrough discovery.

Wen Han, head of research and development of deep potential science and technology macromolecules and researcher of Beijing Institute of Science and Intelligence.

Yuan Huijun, director of the Institute of rare Diseases of West China Hospital of Sichuan University, pointed out that accurate research on rare diseases is becoming the main battlefield for the development of cutting-edge new drugs and life science research tools around the world. The volcano engine Bio-OS bioinformation operating system can help the scientific research of CDGC large-scale deafness cohort projects, the unified management of heterogeneous resources, and the cooperation and convenient interaction of personnel and data in different places.

Yuan Huijun, director of the Institute of rare Diseases, West China Hospital of Sichuan University.

Chen Xin, a researcher at Suzhou Lab, said that with the help of VECTOR, a cloud-based high-performance computing service for volcano engines, Suzhou Lab has effectively supported the requirements of VASP, Gromacs and other material tasks, and achieved remarkable results. Chen Xin said that on the basis of good cooperation, the two sides will continue to deepen cooperation in the field of AI4S material computing.

Chen Xin, a researcher in Suzhou laboratory

The first Bio-OS competition ended, and 15 teams won the prize.

The conference also announced the winners of the first Bio-OS Open Source Competition and held an award ceremony. Xu Tao, academician of the Chinese Academy of Sciences and deputy director of Guangzhou Laboratory, delivered a speech at the ceremony, he said that big data and artificial intelligence have become an important driving force to promote scientific research, Bio-OS Biological big data system as a new bioinformatics analysis support base, its development and application need multiple forces to gather wisdom, common participation, and open source is an important implementation path. The successful holding of this competition is of milestone significance to accelerate the construction of Bio-OS.

Xu Tao, Academician of the Chinese Academy of Sciences and Deputy Director of Guangzhou Laboratory

Sponsored by Guangzhou Laboratory, this competition aims to excavate and train talents in this field, promote the cross-integration of bioscience, data science and artificial intelligence, and contribute to the development of bioinformatics. The competition attracted 187 teams and 217 contestants from 45 universities, 16 scientific research institutions, 8 enterprises and individual participants around the world. After several rounds of judges, a total of 15 teams were shortlisted for the finals and won the prize.

END

With the rapid development of machine learning, deep learning and other AI technologies, artificial intelligence has become a catalyst for scientific research. In the future, volcano engines will also be used for all kinds of scientific research scenarios, through volcano engines to research intelligent clouds, to build AI4S end-to-end solutions from infrastructure to data systems to intelligent applications, and to establish scientific research access networks in multi-dimensional scenarios.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.