Shanghai Jiao Tong University is putting AI computing into research, teaching and campus services, with a university-owned platform rated at 633 petaflops. The project offers a concrete look at what “AI-native” education means beyond the slogan: substantial shared infrastructure, used across campus rather than kept in a lab.
Huawei AI Watch analysis
What happened
The platform, Zhiyuan-1, was developed by Shanghai Jiao Tong University with Huawei. Huawei says it has tens of petabytes of storage and supports large-scale model training as well as research, student tools and university administration.
The reported uses include analysing deep-sea biological samples, coding assistance, language learning and academic support. Huawei also describes a shift towards project-driven teaching built around AI tools and interdisciplinary work. Those are examples of intended and reported campus use, not a measured assessment of educational outcomes.
Why it matters
The interesting part is not simply the headline compute figure. Zhiyuan-1 is being presented as shared university infrastructure, connecting scientific work and everyday campus activity. If that approach proves useful, it could give universities a way to make AI a practical part of research and teaching, rather than a collection of disconnected pilots.
It also raises a question of access: a powerful platform can support a broad campus, but its value depends on who can use it and for what. The announcement describes applications, but does not quantify student uptake or show results against a non-AI comparison.
Our read
This is a substantial campus AI project with a more useful story than “university buys big computer”: the platform is tied to specific research and learning uses. The next test is whether those uses produce visible gains for researchers and students, rather than just an impressive specification sheet. Huawei’s account of the project points to a live experiment in building AI into university life.
What to watch
- Whether the university shares measurable results from research and teaching uses.
- How widely students and staff can access the platform and its AI tools.
- Whether other universities adopt similarly shared, campus-wide AI infrastructure.
Discussion spark: Should universities invest in shared AI platforms for the whole campus, or prioritise smaller tools that can show clear benefits in specific courses and research projects?
Sources and evidence
- Shanghai Jiao Tong University builds an AI-native campus – Huawei (26 October 2026, 08:59 UTC)
not affiliated with or endorsed by Huawei AI