Teaching

Version 11.1 by kindler on 2026/01/26 13:40

BrainScaleS as tool-for-teaching

Both neuromorphic compte systems, SpiNNaker and BrainScaleS, can be used as "tools for teaching" and both systems are accessible also with guest accounts.

BrainScaleS for teaching

The BrainScaleS analog neuromorphic computing hardware systems are operated by the Neuromorphic Computing Architectures, Electronic visions group at the EINC at Heidelberg University in Germany. These hardware systems can be used via Jupyter notebooks (python code) running on the EBRAINS research infrastructure (free of charge for teaching and testing). This access only requires a browser (firefox, chrome, safari, ...), an internet connection and an EBRAINS account (or EBRAINS guest account). Use of the systems in university courses or also secondary school education is explicitly encouraged. Tutorial notebooks with different difficulty levels are available.

  • These tutorial notebooks can be viewed publicly (no account needed) here on github (to get an idea, how they look like see e.g. the single-neuron introduction notebook or the learning with superspike notebook). This is only a static view of the code.
  • With an account the real systems can be used to interactively execute the provided tutorials notebooks and also own experiments:
    • To access a non-persistent, fresh clone of the notebooks use this URL.  This starts a Jupyter notebook session on the EBRAINS research infrastructure, clones the current tutorials from github and opens the index page. The notebooks linked there make use of the real BrainScaleS neuromorphic hardware systems running at the EINC in Heidelberg, Germany. 
      (non-persistent = changes to the notebooks can be made, but they are only stored while the session is running and are gone, when the sessions ends (session timeout is several hours)
    • It is of course also possible to have a place to make persistent changes to the notebooks by creating an own "Collab" in the EBRAINS wiki and cloning the notebooks into that persistent storage. See "Getting the example notebooks into an own Collab" on the quick try page.

To get an own account (requirement: the account owner must be of legal age) use the https://ebrains.eu/register page. If your work email domain is already on the positive list account creation is a matter of minutes (and free of charge).

If the course is intended to be executed by under-aged people, e.g. school kids, we can create "guest accounts", which do not have a minimum-age requirement. These accounts are valid for up to one month (extendable) and can access only parts of the EBRAINS research infrastructure services. BrainScaleS and SpiNNaker systems can be accessed with guest accounts. Please contact the EBRAINS education team, if you need guest accounts for your course plans. 

With the one-click access method to notebooks each student of the course can start her/his instance of a Jupyter session, get clone of the tutorial notebooks and can run and change the notebooks via the EBRAINS RI with real hardware access. (See above for info on the non-persistent and persistent options for changes in the notebooks). Access URLs can go to the index page or directly start a specific notebook, e.g. the single neuron notebook

Tutors can of course create their own set of notebooks by cloning the git repository, changing the clone and then creating a one-click-access-URL to start from that own repository. This can be e.g. done to create notebooks, which contain questions / exercises, but not the answers. Or add notebooks for special topics. (We would be glad to hear about such additional notebooks or any suggestions for improving the existing tutorial notebooks, easiest via the EBRAINS support system by emailing to support@ebrains.eu). 

Lu.i analog neurons for teaching

Please see the Analog Electronic Neuron Demonstrator lu.i page for details about the analog hardware neurons.