3D Gravity Simulations and OpenCL

It has been a few years since I last was experimenting with OpenCL for gravity simulations.

At that time I made some mistakes with my code that resulted in not all of the objects being compared to all of the objects in the gravity calculations. For the purists this is not acceptable “How can you call it a gravity simulation if you are not using all objects in the calculations?!” The other opinion raised at the time was if it looks good enough, use the speed ups. None of the comments to my YouTube gravity videos were (in typical YouTube speak) “haha lol! looks fake! u no do gud gravity!”. So, if you are struggling with getting gravity working fast, try reducing the number of actual particle interactions.

Here is the latest updated CL kernel code….

__kernel void Gravity3DKernel(__global float4* pos, __global float4* vel, __global float4* acc, __global float4* mass, float mingravdist, float forcescalefactor, int startindex, int stopindex, int gridsize)
	int index=get_global_id(0)+startindex;
	float dx,dy,dz,distance,force;
	float positionx=pos[index].x;
	float positiony=pos[index].y;
	float positionz=pos[index].z;
	for(int a=0; a<get_local_size(0); a++) {
		if (a!=index) {

The kernel processes all of the 3d objects once using 1 of the other objects (passed as startindex).

You can call the kernel multiple times each frame/step of the simulation. If you want strictly accurate (and slow) gravity simulation results you call it inside a loop that sets startindex from 1 to the number of objects. In my experiments this is not necessary. If you want “good enough” nice looking gravity simulations then calling the kernel with as little as 10 different startindexes (use a set of objects spread out among all of them or just use a random set of 10 objects each frame).

Here is a recent 4K resolution example movie. This one used 1,000 objects out of 5,000,000 objects for the gravity calculations. The objects start by being randomly placed within an oblate spheroid. Their velocities are initialized so they are rotating around the center axis.


Multiple Rules Cellular Automata

Another new CA to experiment with.

The idea for these came from a comment Tsui Kagura left on one of my YouTube videos.

Can you make a CA that is 4D ( or 5D etc ) in a sense, that you use 2 (or more) different regular CA rules, run them in the same space (same coordinates), same time (same steps), but then, based on some kind of rule also have them interact with each other?As if they were neighborhoods from another dimension. I assume we could only be ‘seeing’ one of them, the rest would be in a hidden dimension, but affecting the one we’re looking at. If there are more than one hidden dimensions, they could affect each other too, maybe using a different rule between them…

Using multiple rules on the same CA is something I have not experimented with before, so I had to give it a go.

To start off I used the simplest 2 state 2D CAs. The extension from a usual 2D CA is simple enough. You run 2 rules over the same grid. Store the results of each rule (either 0 or 1 for dead or alive) and then you use the results of the rules to set the new cell state.

How the 2 rule results are converted into a new cell state is the tricky part.

For two 2D CA rules there are 4 possible outcomes (dead dead, dead alive, alive dead, alive alive). So depending on the result states I have 4 options for alive or dead. This gives the following options.

Multiple Rules Cellular Automaton

Here are a few sample results after trying hundreds of random rules.

Multiple Rules Cellular Automaton

Multiple Rules Cellular Automaton

Multiple Rules Cellular Automaton

Multiple Rules Cellular Automaton

I will update this post if I find any more interesting results. There are many more extensions to try like 3D, 4D, 5D, larger neighborhoods, more than 2 rules, etc.

I did try the 3D version of multiple rules. Nothing worth posting as yet. I used both the result1 and result2 method above and also tried feeding the result of rule1 into rule2. Neither has given any unique looking results yet.