Data Storage Solutions

Here are the different ways to store your data at EPFL

The service desk will be happy to direct you to the right contact person.

Storage at EPFL


Based on the DSI’s file storage service and hosted on the EPFL site, this solution enables collaborative documents to be stored and shared. The service is accessible via the DSI’s XaaS portal, and is open to faculties for the storage of their collaborative administrative data.


The Reseach Computing Platform offers storage for research data. The multi-site, snapshotted (ransomware-resistant), multi-protocol (Mac, Linux, Windows) storage is accessible to labs, schools and platforms. Requests can be made to [email protected]

DSI Object Storage

The object storage service is based on a centralized hosting solution accessible from the entire school IT network and hosted on the EPFL site. The service is accessible via the DSI’s XaaS portal. The main purpose of the service is to enable performance-independent storage of large volumes of data, such as archives or backups. It is not suitable for OLTP applications.

SCITAS Storage

SCITAS provides high bandwidth, low latency storage for scientific computing. Services include medium term storage with 15-day snapshots, multi-site tape backup and long term storage for archiving, as well as export services to Samba, NFS and object storage.

FunctionalityDSI NASRCP NASDSI Object StorageSCITAS Storage
ProtocolsCIFS, SMBNFS, SMBS3CIFS, NFS, S3, Posix
Quota1 TBnonono
Backup Serviceyesnonoyes
Confidential Datayesyesyesyes
Typical UseActive collaboration on Administrative dataActive collaboration on Research dataBackups, Archives, Cold Data – Both adnmin istrative and ResearchScientific data storage and archiving
Price27 CHF/TB/year29 CHF/TB/year25 CHF/TB/year

Cloud Storage

Microsoft 365

This service is hosted in Switzerland by Microsoft. You can use SharePoint Online and OneDrive to securely store, organize, share and access information from any device. All you need is a web browser. It is primarily designed for teams, EPFL units or projects to manage administrative documents. You need to be accredited at EPFL to access it.

Google Workspace

Non-confidential documents only. This service is hosted by Google, and requires an EPFL email address for access. This solution offers a personal (my drive) or team (Shared drive) space for managing scientific and other documents. A quota of 20 GB per user will be introduced in 2023, following Google’s announcement that it will end its unlimited and free offer for Education.


This service is hosted by SWITCH (in Switzerland), and requires a SWITCH edu-ID account for access. This solution offers a personal space. SWITCHDrive can be used to store confidential documents.

FunctionalityGoogle WorkspaceMicrosoft 365SWITCH Drive
Confidential documentsnoyesyes
Group spaceyesyesno
Quota20 Gb250 Gb100 Gb
Search functionyesyesyes
Notification of actionsyesyesyes
Version Managementyesyesno
Web Clientyesyesyes
Permission managementyesyesyes
External access to EPFLyesyesyes
Student accessyesyesyes
Access to a groupyesyesno
Limited access to EPFLyesyesno
Everybody accessyesnoyes

Other solutions available for Research

SCITAS C4Science

c4science is a platform for co-creation, curation and code sharing. This platform includes a version management system, authentication common to all Swiss universities for local and external collaborators, collaboration (wikis, bug tracking, etc.), and a code testing system (continuous integration).

ACOUA Academic Output Archive

ACOUA is reserved for datasets that have led to publications, and requires the signature of a specific agreement for laboratories or projects wishing to start using the service. Data must be prepared for archiving, so remember to plan your requests in advance to make the process as smooth as possible. Datasets can be up to 10 TB in size, with restricted access.

CERN Zenodo

Non-confidential files only. Zenodo is a data repository developed by CERN, allowing the distribution of datasets up to 50 GB, free of charge. Free access is encouraged, but you can also restrict it if necessary. Special quality criteria are required for datasets from the EPFL community.