Research computing

ITSC provides charged research-computing services for CUHK staff. Availability and charges depend on the requested service and available resources.

  • Computing resources — Access research-computing resources for approved work.
  • System planning and implementation — Request planning and implementation support.
  • Software installation — Ask about installing and tuning research software.
  • Advisory services — Discuss requirements with the Research Computing team.

Cloud AI support for research and projects

Researchers and project teams may ask ITSC to assess Microsoft Foundry, Microsoft Copilot Studio, Gemini Enterprise, Amazon Bedrock and Kiro for an approved research, application-development or experimental project.

  • Purpose — Describe the research, development or experimental objective.
  • Users — State who needs access and who is responsible for the project.
  • Data — Describe the information involved and its handling requirements.
  • Platform or models — State any preferred platform, models or required capabilities.
  • Expected usage — Estimate usage and the project period.
  • Funding — Identify the funding source and responsible project owner.

For teaching, learning or pedagogical research, use the AI Services for Teaching and Learning contact instead.

Development access boundaries

  • API tokens are not included with enterprise AI licences.
  • Completing a course or training activity does not guarantee API tokens or quota.
  • API Portal tokens are separate from Copilot Cowork and Perplexity Computer.

Capability separation

API Portal tokens remain distinct from Microsoft 365 Copilot Cowork and Perplexity Computer. All three may be considered under a broader framework for managing usage, budgets and costs, but each is approved, funded, reported and reviewed separately. Quota approved for one is not transferable to another.

Request requirements

The model or service, the purpose, the users, the expected consumption, the access period, the data involved, the development environment, the funding source and the responsible owner.

Get started with the API Portal

The CUHK API Portal for Teaching and Learning supports approved programming classes, teaching experiments, learning activities and proof-of-concept development.

  • Text and conversational models
  • Image-generation models
  • Text-embedding models

Available models, operations, quotas and rate limits vary by API and subscription. Follow the documentation for the API assigned to you.