Gianni – it might be that the initial process is associated with setting up the simulation. When you say “start of the simulation”, do you mean the simulation time is advancing, or are you referring to the “set up” stage, when there is no time integration yet. It might be that memory operations are so overwhelming that the CPU is stalled most of the time waiting for data to move back and forth between CPU and RAM. This is a guess, and if you have 30 mins and get to the bottom of it, it’d be interesting to run a profiler on the code, to see where the time is spent during the dynamics simulation.
One more question: is there a good reason you want to use MCORE? In case you need friction and contact on a large scale, you might take a look at Chrono DEM Engine. Dan --------------------------------------------- Bernard A. and Frances M. Weideman Professor NVIDIA CUDA Fellow Department of Mechanical Engineering Department of Computer Science University of Wisconsin - Madison 4150ME, 1513 University Avenue Madison, WI 53706-1572 608 772 0914 http://sbel.wisc.edu/ http://projectchrono.org/ --------------------------------------------- From: [email protected] <[email protected]> On Behalf Of Gianni Curti Sent: Tuesday, September 24, 2024 5:33 AM To: ProjectChrono <[email protected]> Subject: [chrono] Percentage of CPU decrease over time Hi everyone, I’m encountering some issues with Chrono::Multicore in my simulation, which is based on the demo_MCORE_Cratering example. The goal is to simulate the interaction of an object with a particle bed at rest. The code compiles and runs without errors, but I’ve noticed an unexpected drop in CPU utilization. At the start of the simulation, all the threads I allocate are working at 80-90% capacity. However, as the simulation progresses, their utilization gradually drops to around 10% or even lower. I had expected the computational load to increase as the number of contacts grows from about 7,000 to 50,000, but this CPU behavior is quite strange. I’ve observed this issue on both my Windows laptop (with 20 threads) and a university workstation running Linux (with 40 threads). Has anyone else experienced this? Could it be that the nature of the problem makes parallelization progressively more difficult as the simulation advances? Thanks in advance for your help! Best regards, Gianni -- 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]<mailto:[email protected]>. To view this discussion on the web visit https://groups.google.com/d/msgid/projectchrono/207ae42c-461f-4ce6-b019-cd8a9d8448cfn%40googlegroups.com<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/projectchrono/207ae42c-461f-4ce6-b019-cd8a9d8448cfn*40googlegroups.com?utm_medium=email&utm_source=footer__;JQ!!Mak6IKo!Mn-N_BVwsaeaLOaZsF4s97HOvYjh5IqcOxw7-HDp_xbX7s1PAZ9DDHCyZNSdKhWwJR80RB6Jqjch6ml6kSFF$>. -- 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/DM8PR06MB770369F2A5EDDD76CC486A71B1682%40DM8PR06MB7703.namprd06.prod.outlook.com.
