The Laboratory of Large-Scale Brain Simulation is dedicated to advancing the field of largescale brain modeling.
Our research emphasizes groundbreaking innovations in large-scale brain simulation, focusing on software, algorithms, and computational models.
We are building an ecosystem for brain modeling. You can install it via:
pip install BrainXThis ecosystem is a collection of tools, libraries, and frameworks that can be used to build brain dynamics models and applications.
The ecosystem is designed to be modular, so you can use as much or as little of it as you need. Core components of the ecosystem include:
-
BrainUnit: Comprehensive physical units and unit-aware mathematical system for brain dynamics.
-
BrainCell: Intuitive, parallel, and efficient simulation for biologically detailed brain cell modeling.
-
BrainState: State-based IR compilation for efficient simulation of brain models on CPUs, GPUs, and TPUs.
-
BrainTaichi: The first-generation framework for customizing event-driven operators based on Taichi Lang syntax.
-
BrainEvent: Enabling event-driven computations in brain dynamics.
-
BrainScale: Enabling scalable online learning for brain dynamics:
$O(N)$ complexity for SNNs, and$O(N^2)$ for RNN computations. -
BrainTools: Commonly used tools for brain dynamics programming, for example checkpointing.
For the details of this ecosystem, please see: https://brainmodeling.readthedocs.io/
