pipette
ENEnglish

Bridging In-Situ Visualization to Game Engines: From ParaView Catalyst via ANARI to Unreal Engine

Thomas George, Severin A. T. Klapproth, Jens Henrik G\"obbert, Jonathan Windgassen, Andreas Adelmann, and Sriramkrishnan Muralikrishnan

Preprint

En palabras de los autores

In-situ high-performance computing (HPC) simulations produce massive datasets that remain difficult to stream into game engines for interactive visualization. We present an in-situ, disk-free pipeline that couples simulation data to Unreal Engine 5 through ParaView's ANARI rendering API. A modified ANARI-USD device captures Universal Scene Description (USD) scene graphs in memory at the ANARI layer and streams the engine-agnostic USD format through an embedded ZeroMQ broker to visualization clients, where JuSync, middleware with a companion Unreal Engine 5 plugin, parses the geometry as it arrives and routes it to engine rendering components at runtime, producing scenes that are camera-navigable once received. The decoupled design keeps simulation execution independent of visualization load, and clients may connect or disconnect at any time. The pipeline is demonstrated on the JURECA-DC supercomputer at Forschungszentrum J\"ulich, streaming Free Electron Laser (FEL) simulation data with point clouds of tens of millions of particles.

Resultado principalEl resumen no menciona limitaciones.

Apareció: martes, 20 de octubre. arXiv. Preprint, todavía sin revisión por pares.

Comentario de los autores: 10 pages, 10 figures, 2 tables. Accepted at SC26 (Chicago), ISAV Workshop; to appear in IEEE Xplore, DOI pending