High Performance Particle simulation

Click For Summary

Discussion Overview

The discussion revolves around the challenges and considerations in developing a high-performance real-time particle simulation, specifically targeting a frame rate of 30 frames per second. Participants explore programming languages, mathematical models, and computational techniques relevant to simulating particle behavior influenced by gravitational, pressure, density, and shear stress factors.

Discussion Character

  • Exploratory
  • Technical explanation
  • Debate/contested

Main Points Raised

  • One participant inquires about the recommended programming language for real-time particle simulation, expressing a preference for C++ due to prior experience.
  • Another participant suggests that the choice of programming language is arbitrary and emphasizes the advantages of C++'s object-oriented programming features.
  • The original poster (OP) elaborates on the complexities of simulating dried particles, highlighting the need to account for various physical forces and the high computational cost associated with achieving real-time performance.
  • One participant raises concerns about the feasibility of achieving 30 frames per second in real-time, suggesting precomputation as a potential solution if high-performance hardware is not available.
  • Another participant mentions the availability of a demo n-body graphical simulator that can handle thousands of particles at 30 fps, recommending it for those with an NVidia GPU.
  • There is a suggestion to consider commercial software like Fluent or CFD++ for potentially easier implementation and reduced error rates.

Areas of Agreement / Disagreement

Participants express differing views on the feasibility of achieving 30 fps in real-time simulations, with some suggesting precomputation as a necessary approach. There is no consensus on the best programming language or mathematical model for the simulation, indicating multiple competing perspectives.

Contextual Notes

The discussion does not specify the number of particles involved in the simulation, which may influence the computational requirements and performance outcomes. Additionally, the effectiveness of suggested techniques and models remains unverified within the context of the conversation.

Dido
Messages
4
Reaction score
0
I am looking for a help in finding out the recommended programming language for real time (30 frames/ sec) particle simulation. This simulation involves a lot of calculations and Physics (fluids,...).
I do have a little experience with C++ and I want to concentrate a good part of my time on this simulation.
Please if you know any idea, you are welcome.
I thank you in advance.
 
Technology news on Phys.org
language is arbitrary. You can use any language you are comfortable with. C++'s OOP would make it a little nicer to code, IMHO. You should look up the N-bodies problem, before you really get into it.
 
wow thank you so much.
Actually in my real time dried-particle simulation, I want to deal with gravitational, pressure, density and shear stress (because of shape change due to an applied pressure).
According to my understanding, it have to show a fluidic behavior like when those particles are spread. Within the involved calculations, It is obvious that the computational cost is really high and to achieve the time-sensitive is challenging for me.
Do you know any technic to handle this?
Do you know any mathematical model suitable for this kind of simulation?
Finally, do you know how I can achieve 30 frame/sec for a real time simulation (high performance).
 
Last edited:
FOR UPDATES
---------------
Actually in my real time dried-particle simulation, I want to deal with gravitational, pressure, density and shear stress (because of shape change due to an applied pressure).
According to my understanding, it have to show a fluidic behavior like when those particles are spread. Within the involved calculations, It is obvious that the computational cost is really high and to achieve the time-sensitive is challenging for me.
Do you know any technic to handle this?
Do you know any mathematical model suitable for this kind of simulation?
Finally, do you know how I can achieve 30 frame/sec for a real time simulation (high performance).
 
30fps at runtime is insane. Do you have access The Jaguar? If not, you're going to have to use precomputation. Compute where the stuff should be at every frame, save each as a frame, then use some other program to turn those frames into movies. (or something along those lines)
 
TylerH said:
30fps at runtime is insane. Do you have access The Jaguar? If not, you're going to have to use precomputation. Compute where the stuff should be at every frame, save each as a frame, then use some other program to turn those frames into movies. (or something along those lines)

OP didn't say how many particles.

The http://developer.nvidia.com/cuda-toolkit-32-downloads" includes a demo n-body graphical simulator which does thousands of particles at 30 fps. Download that if you have an NVidia GPU, see how many particles it can simulate for you, and take a look under the hood to see if it can accommodate your problem.
 
Last edited by a moderator:
Can you use a commercial code like fluent or cfd++?
 
I don't see why not. I would be a lot easier and likely less error prone.
 
thx for all your help.
 

Similar threads

  • · Replies 19 ·
Replies
19
Views
4K
  • · Replies 1 ·
Replies
1
Views
3K
Replies
10
Views
5K
Replies
1
Views
2K
  • · Replies 5 ·
Replies
5
Views
1K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 29 ·
Replies
29
Views
8K
  • · Replies 3 ·
Replies
3
Views
6K
  • · Replies 6 ·
Replies
6
Views
3K
Replies
8
Views
2K