We are not having any GPU's in the cluster.
The system requirements in the manual do not mention that a GPU is required:
Minimal configuration
• Windows XP or later (32 or 64 bit), Mac OS X Mountain Lion or later, Debian/Ubuntu with GLIBC
2.13+ (64 bit)
• Intel Core 2 Duo processor or equivalent
• 4 GB of RAM
Recommended configuration
• Windows 7 SP 1 or later (64 bit), Mac OS X Mountain Lion or later, Debian/Ubuntu with GLIBC 2.13+
(64 bit)
• Intel Core i7 processor
• 16 GB of RAM
The number of photos that can be processed by PhotoScan depends on the available RAM and
reconstruction parameters used. Assuming that a single photo resolution is of the order of 10 MPix, 4 GB
RAM is sufficient to make a model based on 30 to 50 photos. 16 GB RAM will allow to process up to
300-400 photographs.
The Photoscan Pro manual also just states that GPU acceleration is supported but it does not state that a GPU is required:
PhotoScan supports accelerated image matching and depth maps reconstruction due to the graphics
hardware (GPU) exploiting.
Regarding your question about OpenGL, we have the following packages (rpm's) installed:
mesa-libEGL.x86_64 17.0.1-6.20170307.el7
mesa-libEGL-devel.x86_64 17.0.1-6.20170307.el7
mesa-libGL.x86_64 17.0.1-6.20170307.el7
mesa-libGL-devel.x86_64 17.0.1-6.20170307.el7
mesa-libGLU.x86_64 9.0.0-4.el7
mesa-libGLU-devel.x86_64 9.0.0-4.el7
Is there any additional rpm required ?