Abstract
Opticks is an open source project that accelerates optical photon simulation by integrating NVIDIA GPU ray tracing, accessed via NVIDIA OptiX, with Geant4 toolkit based simulations. A single NVIDIA Turing architecture GPU has been measured to provide optical photon simulation speedup factors exceeding 1500 times single threaded Geant4 with a full JUNO analytic GPU geometry automatically translated from the Geant4 geometry. Optical physics processes of scattering, absorption, scintillator reemission and boundary processes are implemented within CUDA OptiX programs based on the Geant4 implementations. Wavelength-dependent material and surface properties as well as inverse cumulative distribution functions for reemission are interleaved into GPU textures providing fast interpolated property lookup or wavelength generation. Major recent developments enable Opticks to benefit from ray trace dedicated RT cores available in NVIDIA RTX series GPUs. Results of extensive validation tests are presented.
Highlights
Opticks[1,2,3,4,5] enables Geant4[6,7,8]-based optical photon simulations to benefit from high performance Graphics Processing Units (GPUs) ray tracing made accessible by NVIDIA R OptiXTM[9,10,11]
Recent Opticks developments allow the optical photon simulation performance to benefit from ray trace dedicated processors, called RT cores[13], available in NVIDIA Turing architecture GPUs
Photons are brought to the GPU via NVIDIA OptiX ray generation programs, CUDA ports of photon generation loops and photon generation parameters in the form of "genstep" data structures, which play a central role in the workflow
Summary
Opticks[1,2,3,4,5] enables Geant4[6,7,8]-based optical photon simulations to benefit from high performance GPU ray tracing made accessible by NVIDIA R OptiXTM[9,10,11]. The large size and high photon yield of the scintillator, illustrated, makes the JUNO optical photon simulation extremely computationally challenging with regard to both processing time and memory resources. Opticks eliminates both these bottlenecks by offloading the optical photon simulation to the GPU. Several groups from various neutrino experiments and dark matter search experiments are evaluating Opticks These proceedings give an overview of Opticks, describe its application to the JUNO detector and present performance measurements and validation tests. Further details on NVIDIA OptiX, the use of GPU textures and the CUDA port of Geant photon generation and optical physics are available in the earlier proceedings[5]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have