Small-scale cases: from millions ~ tens of millions of mesh points to billions ~ tens of billions of solution points

Medium-scale cases: from tens of millions ~ billions of mesh points to tens of billions ~ hundreds of billions of solution points

Large-scale cases: starting from billions of mesh points and breaking through to hundreds of billions of solution points

Built on the triangular collaborative architecture of "CPU-GPU high-speed interconnection" and fully asynchronous computing model, supporting scaling up to kilocore supercomputing platforms;

Iteration-free fully asynchronous algorithm enables linear growth of computing cost with simulation scale without convergence issues

High-fidelity large-scale multi-scale system-level simulation with hundreds to hundreds of billions of solution points on low-cost consumer GPU clusters

Integrated solver supporting coupled simulation of fluid, solid, thermal, magnetic and other physical fields, covering subsonic / transonic / supersonic / hypersonic full Mach number range

Flexible combination of multiple surface/volume meshing schemes;

Supports fast fully-quadrilateral mesh generation and re-topology;

Supports multiple volume mesh types, compatible with overset meshes and non-conforming meshes

Deeply optimized data storage and processing with parallel reading capability for ultra-large dataset processing;

Real-time monitoring and post-processing supported via asynchronous output without slowing down simulation computation