Agisoft Metashape supports accelerated image matching, depth maps reconstruction, depth maps based mesh and tiled model generation and photoconsistent mesh refinement operation due to the graphics hardware (GPU) exploiting. 

The processing performance of the GPU device is mainly related to the number of CUDA cores for NVIDIA video chips and the number of shader processor units for AMD and Intel video chips. Additionally depth maps based mesh and tiled model reconstruction as well as photoconsistent mesh refinement operations would benefit from larger amount of VRAM available. 

The table below lists currently supported devices (on Windows platform only). Agisoft will pay particular attention to possible problems with Metashape running on these devices. 

Supported Desktop GPUs on Windows platform 
GeForce RTX 2080 Ti
Radeon VII
Tesla V100
Radeon RX Vega 64
Tesla M60
Radeon RX Vega 56 
Quadro P6000
Radeon Pro WX 7100
Quadro M6000
Radeon RX 580 
FirePro W9100 
 GeForce GTX 1080 Ti Radeon R9 390x 
GeForce GTX TITAN X Radeon R9 290x
GeForce GTX 980 Ti Radeon HD 7970 
GeForce GTX TITAN Radeon HD 6970 
GeForce GTX 780 Radeon HD 6950 
GeForce GTX 680 Radeon HD 6870
GeForce GTX 580 
GeForce GTX 570 
GeForce GTX 560 
GeForce GTX 480 

* CUDA supported devices on Mac OS X may require to install CUDA drivers from official website first for certain MAC OS X versions:  

** Metashape  is  likely  to  be  able  to  utilize  processing  power  of  any  CUDA  enabled  device  with  compute capability 2.0 and higher or OpenCL 1.1 and higher enabled device with SPIR support, provided that CUDA/OpenCL drivers for the device are properly installed. However, because of the large number of various combinations of video chips, driver versions and operating systems, Agisoft is unable to test and guarantee Metashape compatibility with every device and on every platform. However, all GPU-based processing issues should be reported to Agisoft support team for more detailed investigation.

Steps for configuring the GPU:

  • Select Preferences command from the Tools menu.

  • In Preferences dialog select GPU tab.

  • Select available GPU devices in GPU tab of the Preferences window. 

* Use CPU enable flag to allow calculations both on CPU and GPU for GPU-supported tasks. However if at least one powerful discrete GPU is used it is recommended to disable CPU flag for stable and rapid processing. 

** When integrated graphic devices are present it is reasonable to enable them, if no discrete GPU is available, otherwise it is recommended to use only discrete GPU and leave integrated graphics device disabled.

  • Click Apply button. 
  • Click OK button to close the dialog. After that the enabled GPU devices will be used for the GPU-supported processing steps automatically.