The EBRAINS platform provides remote access to two large-scale neuromorphic computing systems: BrainScaleS in Heidelberg and SpiNNaker in Manchester. These hardware resources emulate spiking neural networks using analogue or mixed-signal techniques, offering accelerated simulation speeds up to ten thousand times real-time for studying plasticity and learning. Users can define neuronal network models via the PyNN API through web interfaces or Python clients. The infrastructure supports computational neuroscience and machine learning research by enabling high-speed, low-energy simulations of simplified brain models. Additionally, a printed circuit board demonstrator named 'Lu.i' offers configurable analog neuron implementations for educational purposes.
Use-case examples
- simulation of spiking neural networks
- study of synaptic plasticity and learning
- neurorobotics simulation environments