Center for Advanced Light Microscopy (CALM)

Breadcrumb Navigation



IT strategy at the CALM

Modern microscopes can produce large amounts of data (hundreds of gigabytes to terabytes), the handling which often becomes the bottleneck in imaging-driven scientific projects. Therefore, we provide tools for making data available to internal and external users, analyzing and reducing datasets via advanced processing routines and archiving raw data, while trying to minimize bottlenecks due to e.g. analyses being done on the microscope control workstations.

Microscope workstation configuration 

  • The microscope workstations at the CALM are dedicated to data acquisition
  • Long term storage and data analysis on the acquisition workstations is discouraged
  • The workstations are connected to the BioSysM LAN via Gigabit or 10GiB (spinning disk and lightsheet microscopes) Ethernet

Data transfer

  • Data are transferred from the microscope workstations as soon as possible
  • The use of USB drives is discouraged for security reasons, data are transferred over the network
  • Data can be transferred to internal or external storage servers via SFTP
  • For external users, we also offer to make the data available via the ownCloud server of the biology department Large datasets (10GB-TB) can be temporarily kept on a 100 TB storage server of the CALM
  • In justified cases, we offer storage of raw data on the petabyte tape storage system of the LRZ for 10 years

Data analysis facilities

  • We generally encourage the use of open source analysis tools such as ImageJ/Fiji and Python
  • For small datasets (MB-10GB), users are expected to analyze the data on their own workstations
  • If needed, the imaging workstation of the CALM can be used for this purpose
  • For larger datasets (10GB-TB) and computationally intensive tasks, we have dedicated compute servers at the CALM (one of them containing powerful GPUs for acceleration of tasks like machine learning)
  • Large raw datasets should be archived after analysis, reduced versions can be made available to external via SFTP or ownCloud