Skip to content

gVXR 2.1.0 Release

10 Jun 2026

We are pleased to announce the release of gVXR 2.1.0, the latest version of the gVirtualXRay (gVXR) library, a GPUaccelerated Xray simulation framework for research, education, and industrial applications. 

 

🚀 Availability 

gVXR 2.1.0 is now available on PyPI: 

👉 https://pypi.org/project/gvxr/ 

Supported platforms: 

  • Windows: x86_64 architecture 
  • GNU/Linux: x86_64 and aarch64 architectures 

Install or upgrade: 

pip install --upgrade gvxr 

 

 What’s New in gVXR 2.1.0 

This release builds on 2.0.10 with improvements in performance, usability, and simulation fidelity. 

🔧 Key Updates 

 Mixed-Precision Support 

  • Introduction and refinement of mixedprecision computation 
  • Enables faster simulations by combining half (16 bits) and single (32 bits) precision on GPU where appropriate 
  • Maintains numerical accuracy while improving GPU performance 

 Enhanced Python Integration 

  • Improved PyPI packaging and cross-platform distribution 
  • New digital twin framework to simplify setting up scans for specific devices 
  • Integration with Numpy to speed up the computeXRayImage and makeIsoSurface functions 

🧪 Feature Refinements 

  • Flux can now be set as a number of photons per cm2 at one metre, seamlessly applying the inverse square law when the source-detector distance is changes 
  • Availability of an OpenMP backend on CPU when a GPU is not available 
  • Computation of the peak signal-noise ratio (PSNR) when noise is enabled 
  • Generation of iso-surfaces using the Marching Cubes algorithm to create meshes from 2D images or 3D volumes 
  • Integration of XrayPhysics to generate X-ray tube spectra without any Python dependency 
  • Possibility to automatically maximise the magnification 
  • More consistent behaviour across API functions 
  • Continued maturation of the simulation engine 

🐞 Bug Fixes 

  • Resolution of issues from previous 2.0.x releases 
  • Improved cross-platform compatibility 
  • Fixes for edge cases affecting reproducibility and stability 

👉 Full changelog:
https://sourceforge.net/p/gvirtualxray/gitroot/ci/master/tree/CHANGELOG