External Tools and Services
This section provides a curated overview of external platforms, developer tools, compute services, notebook environments, and AI-assisted development tools that are relevant to AISL teaching, research, experimentation, and project work.
These services are not operated by AISL unless explicitly stated otherwise. They are included as practical references for FHNW users, collaborators, and partners exploring the broader AISL toolstack.
Fast-changing external information: This section is largely assembled through AI-assisted and agentic research workflows and reviewed by humans with relevant expertise. External products change too quickly for every quota, price, model, feature, and eligibility rule to be continuously maintained or manually re-verified. Always check the provider's official documentation before making technical, financial, privacy, or course-planning decisions.
Research and compute infrastructure
CSCS Clariden
Clariden is the CSCS machine-learning cluster on Alps and provides NVIDIA Grace Hopper (GH200) compute for eligible projects, courses, workshops, and research activities.
For FHNW/AISL users with an allocation, the CSCS Clariden quick start covers account verification, signed SSH access, a minimal Slurm test, an optional GPU smoke test, uenv, and the ML Console.
Swiss AI Research Platform / CSCS
The Swiss AI Research Platform is an academic LLM inference service developed as part of the Apertus project. It provides access to selected open models hosted on Swiss research infrastructure.
Best for: Academic experimentation with open models and Swiss AI research infrastructure.
Chat from the FHNW network
The platform also provides its own Open WebUI-based Chat interface. For FHNW users, it is available from the FHNW network, including via FHNW VPN.
API key security
Keep your API key secure. Do not share it publicly or include it in client-side code.
Data collection and use
Request metadata: By using the service, you consent to the collection and use of request metadata. The platform uses this information for research, capacity planning, and service operation.
API prompts and responses: The API does not store prompt or response content. It retains metadata such as user ID, token counts, model, and timestamp.
Chat history: The internal Chat application retains conversation history so users can return to previous chats. Chat content is not extracted and used for model training.
Model training: Prompts and responses are not used to train or fine-tune models.
Non-commercial use only: The service is intended for academic, non-commercial use. Commercial or startup requests should be directed through Apertus contact.
FHNW HPC Lab / Calculon
The FHNW HPC Lab provides high-performance computing infrastructure for demanding applications, particularly in AI, machine learning, data science, and simulation. Its Calculon Cluster supports HPC research and production projects in AI and data science.
Access, allocation, and usage conditions are managed by the FHNW HPC Lab.
Developer, education, and experimentation resources
Before using an external service, verify its current availability, pricing, quotas, privacy practices, model quality, licensing, and terms of use. A service that is free today may change or may not be suitable for confidential or sensitive FHNW data.