Challenges
Insufficient Computing Power
Universities need massive computing power for research, teaching, and management, but face an average shortfall of over 100 petaflops.
Limited Computing Power Sharing
Buying computing power for each research project is inefficient, making it hard to share and fully use resources.
Diverse Computing Needs
Diverse research fields demand varying levels of computing power, necessitating unified resource scheduling tools and policies to meet their specific needs.
Solutions
Scientific Innovation
Scientific innovation acceleration
Teaching Assistance
First-class talent cultivation
Campus Management
Intelligent management
Faculty/Student Services
Refined service experience
Solution Highlights
AI Computing Platform for Education
Efficient and Ease of Use
· Unified compute scheduling and development task management
· End-to-end agile development from data annotation to model deployment
Open Ecosystem
· Compatible with mainstream frameworks and research models
· Wide coverage of MindSpore AI4S suites across electromagnetism, biology, fluid dynamics, chemistry, and materials science
Solid Reliability
· Streamlined resource orchestration across compute, network, storage, and management
· Ascend intelligent cluster O&M engine, ensuring efficient and reliable system running
Related Products
Application Scenarios
Scientific Innovation
Research simulation for more efficient experiment
Academic analysis to align with research directions
Smart document retrieval and research outline creation
Teaching Assistance
Personalized learning and development
Interactive learning for enhanced learner engagement
Automatic exam scoring for streamlined assessment
Campus Management
Proper resource allocation to avoid waste
Financial forecasting to improve fund utilization
Accurate procurement to reduce inventory costs
Faculty/Student Services
AI customer service to improve consultation efficiency
Personalized meal recommendation
AI interview simulation to enhance job readiness
Use Cases
Academy Collaboration
In 2024, a university in Shanghai began building the first Ascend intelligent computing platform equipped with 1,000 cards. Completed and launched in July 2025, the platform delivers peak computing power of 320 petaflops. It now drives advanced research and innovation, supports curriculum upgrades, and fosters interdisciplinary talent—gradually evolving into a globally influential research hub.
Academy Collaboration
In 2024, a university in Nanjing launched an Ascend computing platform with peak power of 85 petaflops to advance AI services and integrate science with education. The platform supports cutting-edge computing research, curriculum development, and talent cultivation. It aims to drive interdisciplinary innovation, nurture top-tier talent, and foster world-class scientific breakthroughs.
Shenyang Institute of Technology
Shenyang Institute of Technology (SIT) has partnered closely with Huawei to integrate industry and education, cultivating digital-era talent through joint initiatives like modern industry colleges, smart campus platforms, and training bases. SIT also built the "Zhiqing·Deep Space" AI computing center on the Ascend platform, supporting model training, optimization, and an intelligent Q&A system for accurate campus services.