his is different than the client server architecture. It means that the server itself was built from the ground up as a parallel application. Everything is thread safe. This means that a single instance of the geometry kernel can answer queries from an unlimited number of meshing threads with no bottlenecks. It means that geometry kernel does not have to be in the same process or even on the same hardware as the mesh generation. Queries are automatically routed to the process that has the data. It means that both the mesh and the geometry can be distributed and the architecture of the server will move geometric queries to the correct process and send the results back. It means that the geometry loading, mesh generation, healing and hundreds of other operations a have been carefully implemented with parallel algorithms.