Nuclideon provides a CPU‑native rendering engine, Unlimited Detail™, that enables real‑time, full‑resolution visualization of terabyte‑scale LiDAR and photogrammetry point clouds on standard x86 CPUs without requiring GPUs. The technology is delivered as on‑premise software (udStream), a secure browser‑based client (udCloud), and an embeddable SDK (udSDK), supporting multi‑format data ingestion, air‑gapped deployment, and enterprise‑grade security.
Funding
Funding not disclosed
Founders
Product
Problem
Large‑scale LiDAR and photogrammetry surveys generate terabyte‑size point clouds, but existing visualization tools require GPU clusters, aggressive down‑sampling, or extensive cloud processing. This forces analysts to work with only a fraction of the captured data, leading to missed details and costly infrastructure. In addition, many regulated environments (defense, mining sites) cannot deploy GPU hardware or rely on external cloud services.
Solution
Nuclideon delivers a CPU‑native rendering engine—Unlimited Detail™—that searches point clouds per‑pixel and streams only the visible voxels to the client. The engine runs on standard x86 CPUs, eliminating the need for dedicated GPUs and allowing full‑resolution, real‑time interaction with datasets of any size. Three product lines expose this capability: udStream for on‑premise or air‑gapped workstations, udCloud for secure browser‑based collaboration, and udSDK as a compiled library that can be embedded in third‑party GIS, C2 or simulation platforms. The technology supports multi‑source data fusion (LiDAR, UAV imagery, CAD, terrain models) and operates across Windows, macOS, Linux and WebAssembly, ensuring consistent performance on laptops, field laptops, or classified servers. Bandwidth usage scales with screen resolution rather than dataset size, enabling low‑bandwidth streaming even over satellite links. All components include role‑based access control, end‑to‑end encryption, and per‑seat licensing that reduces total cost of ownership by up to 80 % versus GPU‑centric stacks.
Target Audience
Primary customers are survey and linear‑infrastructure firms, mining and resources companies, and defense/intelligence agencies that require full‑resolution 3D visualization on standard hardware or air‑gapped networks. Secondary markets include GIS platform vendors and digital‑twin developers seeking an embeddable, GPU‑free rendering layer.
Features
- CPU‑native point‑cloud search algorithm (US Patent 9842425) that renders only visible points each frame, keeping rendering cost constant from 1 GB to 200 TB+ datasets.
- No dedicated GPU required; runs on integrated graphics with a minimum 4‑core CPU, 8 GB RAM, and SSD storage.
- Multi‑format ingestion (LAS/LAZ, E57, PLY, PTS, XYZ, and proprietary scanner outputs) with built‑in conversion tools.
- Air‑gapped and offline deployment options for classified networks, with identical feature sets to cloud deployments.
- udCloud browser client streams via WebAssembly, delivering full‑resolution visualization in Chrome, Firefox, Safari, or Edge with bandwidth proportional to viewport resolution.
- udSDK provides a compiled DLL with a clean C API, supporting integration into Unreal Engine, custom C++/C# applications, GIS platforms, and command‑and‑control systems.
- Collaborative annotation, measurement, and markup synchronization for thousands of concurrent users, backed by patented caching architecture.
- Enterprise‑grade security: TLS encryption, role‑based access control, audit logging, and optional SSO integration.