Organized
by Andrew Davison, Jeff
Muller, Jochen Eppler, David Lester, Abigail Morrison and Markus
Diesmann in collaboration
with the EITN.
Abstract The HBP creates novel
ICT
infrastructure for the neuroscience community. Simulation
engines in software
and hardware for different levels of description are accessible
through a
unified collaboration portal (UP) providing version control and
provenance
tracking. The simulation engine for networks of abstract spiking
neuron models called
the network simulator provided by the software NEST particularly
targets
theoretical neuroscience and neurorobotics. This workshop
presents the
capabilities of the network simulator and the corresponding PyNN
expressions as
available at the end of the ramp-up phase of the HBP in March
2016.
Neuromorphic hardware provided by the NM-PM1 (Heidelberg) and
SpiNNaker (Manchester)
systems enable the translation of a restricted set of software
models to a
platform with accelerated speed and reduced power consumption
for high-throughput
studies and interaction with the real world. With this
background, the main
part of the workshop is reserved for a discussion of the
scientific challenges
and plans of theoretical neuroscience to derive future
requirements on the
simulation engine. This will help developers to optimize the
roadmap of the
second phase of the HBP with the aim to maximize usefulness,
uptake, and
impact.