In February 2024, DataCanvas's self-developed core product, the DataCanvas APS AI Platform, officially joined the National Supercomputing Internet Platform, marking another milestone in the company's commitment to integrated algorithm and computing power services. Responding to the Ministry of Science and Technology's initiative to deploy a national supercomputing internet, this collaboration brings the DataCanvas APS AI Platform to the forefront, especially in scientific research and innovation. It adds AI capabilities to promote integrated supercomputing operations, supporting technological innovation and high-quality digital economic growth.

Efficient Integration of Computing Power and Platform Software

The National Supercomputing Internet, supported and guided by the Ministry of Science and Technology, aims to operate national computing power centers with an internet mindset, striving to build a sustainable and prosperous computing power ecosystem. According to the ministry's plan, by the end of 2025, the National Supercomputing Internet will form an overall layout that is technologically advanced, innovative in its model, excellent in service, and complete in its ecosystem. This will effectively support original scientific innovation, major engineering breakthroughs, and high-quality economic development, becoming the "expressway" for building a digital China.

At the 2023 China National Computing Conference (CCF HPC China 2023), Cao Zhennan, Deputy Director of the National High Performance Computer Engineering Technology Research Center, mentioned, "High-performance computing, intelligent computing, and cloud computing... computing power is showing a trend of diversification; from the user's perspective, they often face the need for multiple types of computing power." Cao also highlighted that generative AI, represented by large models, is flourishing and placing new demands on the "supply side" of computing power.

DataCanvas has been working to provide an efficient computing foundation for the digital economy and offer integrated algorithm and computing power services to the entire market. Fang Lei, Chairman of DataCanvas, stated that computing power construction, as the most crucial infrastructure in the digital economy era, has already begun. It will be a significant industrial revolution, playing an essential role in the progression of civilization. The reserve and operational capabilities of computing resources are essential for the implementation of large-model technologies and intelligent applications. DataCanvas is well-prepared to cooperate with the National Supercomputing Internet, leveraging its AI Foundation Software (AIFS) products, to provide high-quality AI foundational services for scientific innovation and industry breakthroughs.

As a leader in AI foundational software, DataCanvas efficiently integrates computing power and platform software. By participating in the construction of intelligent computing centers, DataCanvas builds an AI infrastructure system that includes computing power, algorithms, and data, achieving integrated "algorithm + computing power" services. This comprehensive approach serves global industry leaders and Fortune 500 companies across various sectors such as finance, telecommunications, manufacturing, energy, transportation, and aviation.

From Experiment to Practice: Enabling AI Applications

Just as AI methods have boosted scientific research across various fields, making the potential of AI for Science a consensus, different subfields are exploring AI-driven scientific research. The true integration of AI methods with high-performance, flexible computing power services can help researchers more efficiently and conveniently conduct specialized research in the "platform science" field. However, this comprehensive integration still has a way to go.

DataCanvas has long been dedicated to the field of AI foundational software, independently developing the AIFS suite. This includes a "general + industry" open-box multimodal large-model matrix, an AI platform supporting full-process development and application of large and small models, an automatic business modeling platform, causal learning software, and a large-model application ecosystem. These all-open, highly automated, and highly collaborative software tools support users throughout the entire model construction to application process.

The DataCanvas APS AI Platform, a core component of the AIFS suite, has now been launched on the National Supercomputing Internet. The platform helps researchers, industry experts, and data scientists handle multi-source, heterogeneous data and quickly, efficiently develop, train, and deploy machine learning and deep learning models of any scale. This facilitates the transition of AI applications from experimental to practical stages, bridging the last mile of industrial-grade "large + small" model applications.