Models	Platforms	Techniques
Brian (2.1)	Point and multicompartmental neurons; plastic and static synapse models	CPUs; GPUs via GeNN	AST transformations; Symbolic model analysis; Code optimization
GeNN (2.2)	Models that can be defined by timestep update code snippet; mostly point neurons and synapses with local update rules	GPUs and CPUs	Direct code generation by a C++ program
Myriad (2.3)	Compartmental neurons; arbitrary synapse models	CPUs; GPUs	Custom object models; AST transformations
NESTML (2.4)	Point neurons	CPUs via NEST	Custom grammar definitions; AST transformations; model equation analysis
NeuroML/LEMS (2.5)	Point and multicompartmental neurons; plastic and static synapse models	CPUs via NEURON and Brian; SBML	Procedural generation; template-based generation; semantic model construction
NineML (2.6)	Models defined by a hybrid dynamical system; mostly point neurons and synapses with local update rules	CPUs via NEURON, NEST and PyNN	symbolic analysis; template-based generation
NEURON/NMODL (2.7)	Point and multicompartmental neurons; plastic and static synapse models; linear circuits; reaction-diffusion; extracellular fields; spike and gap junction coupled networks	CPUs; GPUs via CoreNEURON	Custom grammar; parse tree transformations; GUI Forms
SpineML (2.8)	Models defined by a timestep update code snippet; mostly point neurons and synapses with local update rules; generic inputs support compartments and non-spiking components	CPU via BRAHMS and PyNN; GPU via GeNN and Neuorkernel	XSLT code templates and libSpineML
SpiNNaker (2.9)	Common point neuron models with either static of plastic synapses	SpiNNaker	Hand crafted modular source code, loaded through a complex mapping process from a graph representation
TVB-HPC (2.10)	Neural mass models	CPUs; GPUs	AST transformations