Overview#
An analyst investigating a complex financial crime network pulls up 80,000 nodes in the graph viewer. On a standard laptop with integrated graphics, the force-directed layout stutters, the VR render is unworkable, and the 3D terrain map tiles load slowly. On a workstation with a discrete GPU and sufficient VRAM, the same dataset runs smoothly in stereo VR at 90 Hz, the analyst can spatially navigate relationships in three dimensions, and terrain overlays render without degradation.
Argus supports immersive exploration of large relationship graphs using WebXR (VR/AR in the browser) and GPU-accelerated 3D rendering. For customers working with larger graphs (tens of thousands of visible nodes) or enabling 3D terrain maps via CesiumJS and MapLibre GL, the difference between a fluid experience and motion-sickness-inducing stutter comes down to a small number of hardware factors. This page provides tiered recommendations for procurement and deployment planning.
Open Standards#
- W3C WebXR Device API: The immersive VR and AR graph exploration runs entirely in the browser via the WebXR Device API, enabling headset discovery, session management, and stereo rendering without native plugins.
- WebGL 2.0 (Khronos Group): GPU-accelerated rendering of large graph layouts and 3D terrain overlays relies on WebGL 2.0, which provides the shader pipeline and instanced-draw capabilities needed for high-node-count scenes at 90 Hz.
- OGC 3D Tiles 1.0: CesiumJS consumes and serves geospatial 3D content in the OGC 3D Tiles streaming format, allowing terrain tilesets to be loaded progressively based on viewer frustum and level of detail.
- GeoJSON (RFC 7946): Geospatial features overlaid on 3D terrain maps are exchanged as GeoJSON FeatureCollections, providing a common wire format between the graph backend and the CesiumJS visualisation layer.
- GEXF 1.3 (Graph Exchange XML Format): Relationship graph data can be exported in GEXF 1.3 for interoperability with external graph analysis tools such as Gephi, preserving node properties, edge weights, and visual colour attributes.
- GraphQL: All graph topology, structural-scoring, and CesiumJS tileset data are queried and mutated through a GraphQL API, enabling clients to request precisely the subgraph or asset metadata required for a given VR scene.
- OAuth 2.0 (RFC 6749): Access to 3D tileset endpoints on Cesium ion and compatible servers is authenticated using OAuth 2.0 Bearer tokens, controlling which tilesets a given session may load.
Last Reviewed: 2026-02-04 Last Updated: 2026-04-14
Key Features#
What Drives Performance#
Large graphs stress the system in two main ways. VR renders the scene twice (one view per eye) at high refresh rates, commonly 72 to 90 Hz. Graphs can include many visible nodes and edges at once, and some views run layout simulation (force-directed motion) and spatial queries. Even if the GPU is strong, a weak CPU can cause frame pacing issues.
3D terrain maps in Argus use MapLibre GL with raster DEM terrain tiles, terrain exaggeration, optional hillshade and contour lines, and CesiumJS for advanced 3D visualisation. These workloads benefit from both GPU VRAM and reliable WebGL2 support.
Recommended Hardware Tiers#
The tiers below are designed for customer procurement conversations. "Graph scale" is practical guidance, not a hard limit. Edge density (how many connections per node) can cost more than node count.
| Tier | GPU | RAM | Use Case |
|---|---|---|---|
| 1: Medium | Discrete GPU, 8 GB VRAM | 16 GB | Medium graphs, standalone VR, 3D terrain |
| 2: Large | Discrete GPU, 12 GB VRAM | 32 GB | Large graphs, PCVR streaming |
| 3: Enterprise | Discrete GPU, 16 GB+ VRAM | 64 GB | Very large graphs, high-resolution headsets |
GPU Selection Notes#
- A discrete GPU is strongly recommended for large graphs and for any VR use.
- Favor GPUs with 12 GB or more VRAM when planning for large graphs and high-resolution headsets.
- If choosing between CPU and GPU upgrades, upgrade the GPU first for VR comfort.
Headset Guidance#
Because Argus uses WebXR for immersive graph viewing, headset support depends on the headset browser and runtime.
Recommended default: Meta Quest 3. The Quest 3 is a strong default for customers because it supports WebXR in-headset for quick standalone usage, supports hand tracking on compatible experiences, and can scale up via PCVR streaming when graphs become very large.
PCVR-first headsets suit customers who prefer higher tracking fidelity or different ergonomics. If you go this route, validate the full chain: VR runtime, browser support for WebXR, GPU drivers, and deployment environment.
Connection modes for Quest 3:
- Wired (USB-C Link): most consistent performance
- Wireless streaming: requires strong Wi-Fi and clean RF environment; prefer Wi-Fi 6/6E with good signal strength
Use Cases#
- Public safety agencies requiring comprehensive operational tools with graph-based investigation views
- Organizations managing complex multi-stakeholder networks requiring spatial relationship exploration
- Teams requiring real-time data analysis across large connected datasets
- Intelligence analysis workflows where spatial navigation surfaces non-obvious relationships
- Multi-agency operations requiring coordinated information sharing across complex entity graphs
Integration#
- WebXR runtime integration for immersive VR and AR graph exploration
- GPU-accelerated WebGL2 rendering for large graph and terrain workloads
- CesiumJS 3D terrain visualisation for geospatial graph overlays
- Force-directed layout simulation integrated with spatial query engine