Federate your data.
Deliver it anywhere.

The Open Science Data Federation connects your existing storage to a national network of caches — so data lands next to the compute, without copies or migrations.

Origin

Bring your own data source.

Point an Origin at storage you already run. FabAID exposes it to the federation through Pelican — no re-hosting, no lock-in.

S3 & object stores — AWS, Ceph, MinIO and S3-compatible endpoints.
Globus & POSIX — campus filesystems and existing Globus collections.
Access controls preserved — your authorization rules travel with the data.

31 OSDF caches

Live across the federation

How it works

From your repository to the cache, in three hops.

STEP 01

Register an Origin

Connect your storage endpoint. Pelican advertises your namespace to the federation.

STEP 02

Caches pull on demand

Caches near most US institutions fetch and hold hot objects close to compute.

STEP 03

Workloads read locally

Jobs at any Access Point read from the nearest cache at line rate — no manual staging.

On the fabric now

Real science moving data over the OSDF.

See all 176 OSDF projects

OSDF project

PlantPathology_Gluck-Thaler

PI: Emile Gluck-Thaler · University of Wisconsin–Madison

Biological and biomedical sciences

Our mission is to understand the origins and outcomes of fungal interactions that threaten plant and human health.

13.1 PB

Data delivered over the OSDF

3,498,387

Jobs

10.4M

Files via OSDF

314.6K

CPU hours

0.2

GPU hours

View this project

Built on Pelican

An open platform for federating data repositories.

Pelican ↗

OSDF caches

Namespace registry

Get involved

Bring your data onto the fabric.

Request an access point and connect your first repository in an afternoon — facilitation is free.

Federate your data.Deliver it anywhere.