CoreNEURON

CoreNEURON is a compute engine for the NEURON simulator optimised for both memory usage and computational speed on modern CPU/GPU architectures. Some of its key goals are to:

• Efficiently simulate large network models
• Reduce memory usage
• Support execution on specialised hardware, GPUs
• Support optimisations such as vectorisation and cache-efficient memory layout (e.g. Structure-Of-Arrays)

CoreNEURON is tightly integrated with NEURON, and can transparently simulate a large number of NEURON models. Certain conditions that must be met are covered in the CoreNEURON Compatibility section.

• CoreNEURON Compatibility
  • Unsupported Features
  • MOD File Compatibility
    • Thread Safe MOD Files
    • TABLE Usage With GPU Execution
    • NEURON Only MOD Files
    • Explicit ION Variables Update
    • Random Number Generators: Random123 vs MCellRan4
    • Memory Management for POINTER Variables
    • Have Question?
• Getting CoreNEURON
  • Installing with pip
  • Installing from source
    • Compiler Selection
    • Step-by-step instructions
• Running a simulation
  • Building MOD files
  • Enabling CoreNEURON
• Examples