Hello Yan, Thank you for pointing this out and bringing it to our attention. We are working on optimizing the LiDAR point cloud save option and will update this thread when it's done.Thank you.
Best, Nevindu M. Batagoda On Wednesday, October 23, 2024 at 8:46:30 AM UTC-5 Yan Dong wrote: > > Hi, I check the running, and I find "LiDAR Output" seems to be the lowest > part. > Without > "PushFilter(chrono_types::make_shared<ChFilterSavePtCloud>()" > line, the simulation is about 4X faster. > > However, I have to save the output. > I hope this engine can optimized so that the lidar output uses a new > thread with a buffer, so that the simulation bottleneck is not the IO. > > Thanks. > 在2024年10月22日星期二 UTC+8 17:42:44<Yan Dong> 写道： > >> Hi, >> >> I simulate a moving LiDAR on a user-defined platform which only has >> severl ChEasyBody and springs, >> The environment is a large mesh obj which is about 500MB, and I only save >> lidar's scan, >> No collision is detected. >> >> It take several miniutes to run a one-miniute trajectory. >> Release mode, C++. Compiled by Visual Studio, on Windows10 system. Chrono >> version 8.0.0 >> >> Our computer is a high-end server, which has powerful CPUs and GPUs, >> Memory and SSD Disk. >> We notice that when running the simulation, CPU usage is <5% and GPS is >> <20%. So did Memory and Disk. >> We tested our hardware and it shows that there's no issue with CPU, GPU, >> Memory and Disk. >> >> It seems that the program does not work efficiently. My questions are: >> 1) Is it reasonable that such simple the simulation is so slow? >> 2) and, and suggestions to improve? >> >> Thanks. >> Larry >> > -- You received this message because you are subscribed to the Google Groups "ProjectChrono" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/projectchrono/8d1273a4-6d4c-4564-b7e2-7ef994d9bedbn%40googlegroups.com.
