EDEN: A High-Performance, General-Purpose, NeuroML-Based Neural Simulator

Area of Science:

• Computational neuroscience
• Neuroscience modeling
• Scientific simulation

Background:

• Modern neuroscience relies on detailed in silico neural network models.
• High computational performance, reproducibility, and transparency are crucial for these models.
• Existing simulators may require learning new languages and lack optimal performance.

Purpose of the Study:

• To present EDEN (Extensible Dynamics Engine for Networks), a novel neural simulator.
• To achieve high model flexibility and computational performance for in silico neuroscience.
• To enable direct use of NeuroML-v2 models without additional user effort.

Main Methods:

• Developed EDEN, a general-purpose, NeuroML-based neural simulator.
• Utilized an innovative model-analysis and code-generation technique.
• Validated EDEN using models from NeuroML-DB and Open Source Brain repositories.

Main Results:

• EDEN directly runs NeuroML-v2 models, enhancing usability.
• EDEN demonstrated functional correctness against the NEURON simulator.
• EDEN achieved 1-2 orders of magnitude performance improvement over NEURON on standard hardware.
• EDEN is designed for seamless scaling across CPUs and computer clusters.

Conclusions:

• EDEN offers a significant advancement in neural simulation performance and flexibility.
• The simulator supports the growing demand for detailed and reproducible in silico neuroscience.
• EDEN facilitates efficient large-scale neural network simulations.