Use of CUDA architecture
Posted: 18 Oct 2016, 14:59
Hello,
The simulation I am working on uses several vision sensors such as LIDAR, stereo cameras, etc, which makes the simulation slow.
Recently, I have read about the parallel computing capability of many NVIDIA cards, CUDA. V-REP does not seem to make use of this feature, since I have CUDA installed and my GPU utilization barely reaches a 30% at maximum demand.
Apparently, in other applications, the use of this technology results in a drastic increment of the performance, so it might be a good way to handle vision related calculations.
Could V-REP make use of this technology? Is V-REP expected to fully support these kind of graphic cards?
Regards
The simulation I am working on uses several vision sensors such as LIDAR, stereo cameras, etc, which makes the simulation slow.
Recently, I have read about the parallel computing capability of many NVIDIA cards, CUDA. V-REP does not seem to make use of this feature, since I have CUDA installed and my GPU utilization barely reaches a 30% at maximum demand.
Apparently, in other applications, the use of this technology results in a drastic increment of the performance, so it might be a good way to handle vision related calculations.
Could V-REP make use of this technology? Is V-REP expected to fully support these kind of graphic cards?
Regards