This article provides information on the GPU's effect on processing speed, the minimum requirement for a GPU, recommended GPU, and how to enable GPU processing in PIX4Dmapper.
- For more information about Hardware Components usage when processing with PIX4Dmapper: Hardware components usage when processing with PIX4Dmapper.
- For more information regarding: Mac / Windows XPsg / Linux / Remote Access - Virtual machine / Distributed - Parallel processing: macOS / Windows XP / Linux / Remote access / Parallel processing.
- For more information about the processing speed: Processing speed.
- For more information about updating GPU driver: How to update the driver for an NVIDIA graphics card - PIX4Dmapper
PIX4Dmapper leverages the power of Nvidia GPUs for faster processing.
The speed increase depends highly on the project, the number of images, the image size, and the image content. For example, using a GTX 970, speed up between 10% and 75% on step 1. Initial Processing, and around 10% in step 2. Point Cloud and Mesh were observed. As a rule of thumb, projects with high overlap, high image content, and a high number of key points can be benefited more from the speedup.
Minimum requirements for CUDA speedup
Nvidia GPU (GTX, Mobile, Titan, Quadro) with at least 2GB RAM, supporting CUDA 9.1 with the latest driver installed (starting at least from 343).
Recommended GPU
GTX 750, GTX 970 (around 200-400$), GTX 1060, 1070, 1080 with the latest driver installed.
How to enable GPU processing
If a GPU with CUDA support is automatically detected, it will be used for processing automatically.
The use of the GPU can be enabled or disabled in the Processing Options:
1. In the Menu Bar, click Process > Processing Options...
2. Select Resources and Notifications.
3. In the Maximum Resources Available for Processing section, select or unselect the GPU device.
- When having a CUDA compatible GPU, it is recommended to verify that the GPU is detected by the software in the Processing Options.
- Devices with Compute Capability lower than 2.0 are not supported by PIX4Dmapper.
- For more information about how to check the Compute Capability: http://en.wikipedia.org/wiki/CUDA#Supported_GPUs.
- For more information about the Computer Capability: https://developer.nvidia.com/cuda-gpus.