100,000 Particles at 30fps on Quad Core i7

Loading...

Sign in or sign up now!
Alert icon
Upgrade to the latest Flash Player for improved playback performance. Upgrade now or more info.
2,032
Loading...
Alert icon
Sign in or sign up now!
Alert icon

Uploaded by on Aug 16, 2011

I've just gotten a MacBook Pro with 2.0 GHz Quad Core i7. The MPM simulator is scaling very nicely. Video capture slowed it down a bit and I've sped it up to 30fps to show what it's like without it.

Category:

Gaming

Tags:

License:

Standard YouTube License

Link to this comment:

Share to:

Uploader Comments (kotsoft)

  • Really, really nice. What Poisson solver do you use for pressure ? Is it simple Jacobi ? If so, try SOR, it's just a few more lines, but it really improves convergence, and you will have less elastic effect. I had also tried conjugate gradient, but way to complicated (I like simple code ^^), and sometimes troubles at boudary conditions. My experience is that SOR is the most valuable solver for real time simulation.

  • @VelRyphon i'm just using a non-iterative pressure. it is already mostly pretty stiff and compressible only when large regions are moving quite a bit faster than the cell size, which normal incompressible free surface simulations have trouble with anyway. also, at 100,000 particles this is a pretty large scale simulation so if the room's a little more irregular with smaller sections and more zoomed in, the fluid will feel pretty incompressible. also, i kind of like the shockwaves.

  • I've download Polluted Plantet, it runs smoothly on my Phenom II x4 955 3.6 GHz CPU. Very nice

  • @magleman hey, could you share percentages for each core? i'm just curious for scaling purposes.

  • @kotsoft It uses all 4 cores equally, about 70%

  • @magleman ok thanks. it seems that either memory or gpu is being a bottleneck. i'm going to be experimenting a little with cache coherency to see if i can get rid of that bottleneck. what gpu do you have?

see all

All Comments (30)

Sign In or Sign Up now to post a comment!
  • Holy crap, this is incredible! Is this open-source? CS student here!

  • @kotsoft ATI Radeon HD5870

  • @kotsoft Ok. Really nice implementation, anyway. Have you tried to simulate a drop splash ? The worthington jet is the most difficult thing to simulate for my own solver (SPH one). I'm curious about how MP handles this test case.

  • @Rezelmombah i'll release one when i have a bit more features such as different tools and i need some interface to edit material properties and stuff.

  • @kotsoft when will you release a demo?

Loading...

Alert icon
0 / 00Unsaved Playlist Return to active list
    1. Your queue is empty. Add videos to your queue using this button:
      or sign in to load a different list.
    Loading...Loading...Saving...
    • Clear all videos from this list
    • Learn more