Cloud-Based Photogrammetry Benchmarking

Digital Twins Geospatial Industry Nebula Cloud HPC Photogrammetry NVIDIA GPU

Photogrammetry processing in the cloud is revolutionizing industries such as geospatial mapping, heritage conservation, construction, and digital twin creation. However, selecting the right cloud-based infrastructure is crucial for optimizing performance and cost. This blog provides insights into benchmarking photogrammetry software on Nebula Cloud instances, focusing on Agisoft Metashape, Pix4D, and RealityCapture with both CPU and GPU-based processing.

Why Cloud-Based Photogrammetry?

Unlike on-premise setups, cloud-based photogrammetry offers:

• Scalability: Process large datasets with on-demand resource allocation.
• Cost Efficiency: Pay only for the computing power you need.
• Performance Optimization: Leverage high-end GPUs and CPUs for faster processing.
• Global Accessibility: Collaborate remotely with cloud-based storage and processing.
• Seamless Processing of Large Datasets: Nebula Cloud Workbench can process terabytes of satellite and drone imagery efficiently.
• Digital Twin Production: Supports NeRF and Gaussian Splatting techniques for high-accuracy 3D reconstructions and immersive visualization.
•All-in-One Photogrammetry Solution: Provides seamless access to all popular photogrammetry software in one place, enabling users to select the best tool for their projects based on workload and processing requirements.

Nebula Cloud Instance Recommendations for Photogrammetry Processing

1. CPU-Based Processing (For Batch Processing & Cost Efficiency)

Recommended for workflows that require high compute power but do not necessarily need GPU acceleration.

Optimizations:

• Enable multithreading in Agisoft Metashape and Pix4D.
• Increase cache size and RAM allocation.
• Use fast SSD storage for temporary files.

2. GPU-Based Processing (For Faster Results & Large Datasets)

GPU acceleration significantly speeds up image alignment, dense point cloud generation, and mesh reconstruction.

Optimizations:

• Enable CUDA acceleration in software settings.
• Assign more VRAM for texture and depth map generation.
• Store datasets on NVMe SSDs for faster access.

Software Tuning Recommendations for Agisoft Metashape, Pix4D, and RealityCapture on Nebula Cloud:

Optimizing Agisoft Metashape, Pix4D, and RealityCapture requires specific tuning for CPU, GPU, memory, and storage settings. Below are the best configurations for each software:

Key Features of Nebula Cloud Workbench for Photogrammetry:

- Intelligent Task Splitting: Users can now split CPU and GPU-based activities for optimal utilization and cost efficiency. This unique feature is only available on Nebula Cloud Workbench.
- Flexible Compute Allocation: Run CPU-intensive tasks on memory-optimized/HPC instances while executing GPU-intensive tasks on a variety of GPU instances depending on the geoprocessing activity.
- Dynamic Scaling: Users can switch between CPU and GPU instance types and scale up or down dynamically based on their workload demands.
- Optimized Cost Management: Ensures efficient resource utilization, preventing unnecessary costs while maintaining high performance.
- Seamless Large Dataset Processing: Nebula Cloud Workbench supports processing terabytes of satellite and drone imagery seamlessly.
- Advanced Digital Twin Production: Supports NeRF (Neural Radiance Fields) and Gaussian Splatting for high-fidelity 3D model creation and interactive visualization.
- Comprehensive Software Access: Nebula Cloud Workbench provides seamless access to all popular photogrammetry software, allowing users to select the best tool for their projects based on the type of workload and required processing efficiency.

Selecting the right Nebula Cloud instance for photogrammetry depends on your dataset size, processing speed requirements, and budget. Here’s a quick takeaway:

- For budget-conscious users: Start with Tesla T4 (GPU-Optimized-T4).
- For high-speed processing: Use GPU-Optimized-A10 with A10G GPUs.
- For enterprise-scale tasks: Opt for GPU-Optimized-A100 with A100 GPUs.

By leveraging cloud-based computing, photogrammetry professionals can achieve faster results, reduce costs, and scale seamlessly. Have you tried cloud-based photogrammetry? Share your experiences in the comments!

About Nebula Cloud Workbench for Photogrammetry

Nebula Cloud offers a ready-to-use Photogrammetry Workbench with pre-configured environments for Agisoft Metashape, Pix4D, RealityCapture, and more. It also supports cutting-edge digital twin production using NeRF and Gaussian Splatting for next-generation 3D modeling. Additionally, users get seamless access to all major photogrammetry software, ensuring they can select the best tool for their projects. Try it today!