The laboratory runs its own computing infrastructure for training and serving artificial intelligence models, for data analysis and for research data management, and its members have access to professional analytics and cloud platforms.
Computing infrastructure
140 TB
Total storage capacity, 128 TB on Synology network storage and 12 TB on local SSDs
×3Apple Mac Studio M3 Ultra
Each with 256 GB of unified memory and a 4 TB SSD, for training and running models with Metal acceleration and for computationally demanding analysis in Python and R.
×2NVIDIA RTX 5090 GPU servers
Servers with NVIDIA GeForce RTX 5090 cards for deep learning training in a CUDA environment.
×4Synology storage units
Network storage of 32 TB per unit, for research data, backups and shared datasets across the members of the laboratory.Software and platforms
Members of the laboratory have access to the following analytics and cloud platforms.
Palantir Foundry
Databricks
Microsoft Azure
Amazon Web Services
Posit (RStudio)
Anthropic
DataCamp
OpenAI
Google Cloud