Yes, the CPU's single thread performance matters in your case. 1.66 RTF is actually not as bad as I'm thinking. I am suspecting you are using a terrain obj mesh which is way too detailed. As the question I proposed in the last answer, could you please post the number of triangles and the size of the terrain?
I suggest to use Meshlab or Blender to simplified the rigid terrain profile. Another option can be slicing the terrain obj into several pieces, such as slicing a squared obj terrain into four sub-terrains. Jason On Monday, August 29, 2022 at 10:25:26 PM UTC-5 stefan wrote: > The time step is 0.5e-3, It may be a little small, but it can give a > reasonable result. I am using the sedan vehicle and the obj file is input > as rigid terrain. I don't know how to change the number of triangles and > terrain size, could you give me some advice about how to speed up my > simulation by changing the terrain size and number of triangles? A > separate rendering engine, such as Carla UE, was used and it run on a > different computer. I have run my simulation on two different computers, > and the cpu usage both are 100%. The RTF are 2.5 and 1.66 respectively. It > depends on the performance of the single cpu core. > 在2022年8月30日星期二 UTC+8 01:51:34<JASON Z> 写道: > >> Stefan, >> >> Theoretically, you should be able to run one vehicle per thread on rigid >> terrain using most of modern machines. If you are performing vehicle >> simulations with more than one vehicle, you can look into Synchrono. >> >> >> Many parameters can affect the RTF of the simulation. Could you please >> provide more information regarding the simulation you are running? Time >> step, what vehicle model, terrain size, number of triangles you used for >> terrain mesh? What rendering engine you are using, and what RTF are you >> currently getting? Maybe a screenshot can help as well. >> >> Thanks, >> Jason >> >> On Monday, August 29, 2022 at 7:02:55 AM UTC-5 stefan wrote: >> >>> Hi marcel, >>> >>> The real-time simulation is exactly what I want. I have already >>> decoupled physics from graphics, but it still can not meet my requirements. >>> Is it possible to >>> separate the tire model and vehicle model? In that way, the computation >>> of tire and vehicle can be run on separate cores. >>> 在2022年8月29日星期一 UTC+8 17:48:53<[email protected]> 写道: >>> >>>> Hello Stefan, >>>> >>>> I guess you want to try to speed things up to ensure that you can run >>>> your simulation in real-time? As far as I know there is no way to >>>> dramatically speed up the simulation at the moment. You need a relatively >>>> small time step to keep it all stable and there is no easy way to run >>>> those >>>> calculations in parallel. The only thing you can do to speed things up is >>>> to decouple physics from graphics so they run on separate cores. >>>> >>>> Greetings, Marcel >>>> >>>> >>>> On 8/29/2022 8:48, stefan wrote: >>>> >>>> Is there any way to speed up the chrono vehicle simulation with sedan >>>> vehicle in Tmeasy tire or pacejka tire on rigid terrain road? >>>> I found two modules in chrono, One is syschrono,and the other one is >>>> multicore. >>>> >>>> As far as I am concerned, the synchrony is for autonomous vehicles and >>>> can decrease the simulation with multiple vehicles. And the other one is a >>>> multicore module, it can not speed up the simulation >>>> >>>> -- >>>> 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 on the web visit >>>> https://groups.google.com/d/msgid/projectchrono/567a453d-4781-485b-bfba-69ab2f9135b1n%40googlegroups.com >>>> >>>> <https://groups.google.com/d/msgid/projectchrono/567a453d-4781-485b-bfba-69ab2f9135b1n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> >>>> -- 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 on the web visit https://groups.google.com/d/msgid/projectchrono/e92ffbba-d186-4230-872a-09a056e133fbn%40googlegroups.com.
