I have millions of objects that I want to render. I found that instantiate is not suitable, so I'm using Graphic.RenderMesh which is amazingly fast and light. But here comes the problem: it takes CPU resources because we are using a loop for each object and then rendering it:

void DrwaTree(Vector3 position, Quaternion rotation, Vector3 scale)
{
    Matrix4x4 matrix = Matrix4x4.TRS(position, rotation, scale);
      
    for (int i = 0; i < materials.Length; i++)
    {
        RenderParams rp = new RenderParams(materials[i]);
        rp.shadowCastingMode = TreeShadowCasting[i];
        Graphics.RenderMesh(rp, TreeMesh, i, matrix);
    }
}

This function is called a million times in different scripts with different meshes. I want to find a way to make it less CPU-intensive.

One current approach I am applying is to bring different objects under a common center point (multiple center points for different objects). Then, I calculate the distance between the center point and the camera position to execute a specific loop only for objects near the camera. Is there any alternative or more appropriate approach?

I have millions of objects that I want to render. I found that instantiate is not suitable, so I'm using Graphics.RenderMesh which is amazingly fast and light. But here comes the problem: it takes CPU resources because we are using a loop for each object and then rendering it:

void DrwaTree(Vector3 position, Quaternion rotation, Vector3 scale)
{
    Matrix4x4 matrix = Matrix4x4.TRS(position, rotation, scale);
      
    for (int i = 0; i < materials.Length; i++)
    {
        RenderParams rp = new RenderParams(materials[i]);
        rp.shadowCastingMode = TreeShadowCasting[i];
        Graphics.RenderMesh(rp, TreeMesh, i, matrix);
    }
}

This function is called a million times in different scripts with different meshes. I want to find a way to make it less CPU-intensive.

One current approach I am applying is to bring different objects under a common center point (multiple center points for different objects). Then, I calculate the distance between the center point and the camera position to execute a specific loop only for objects near the camera. Is there any alternative or more appropriate approach?

I have millions of objects that I want to render, found that instantaite is not fit so using Graphic.RenderMesh which is amazingly fast and light. But here comes the problem that it takes CPU resources cause we are using a loop for each object and then rendering:

void DrwaTree(Vector3 position, Quaternion rotation, Vector3 scale) { Matrix4x4 matrix = Matrix4x4.TRS(position, rotation, scale);

    for (int i = 0; i < materials.Length; i++)
    {
        RenderParams rp = new RenderParams(materials[i]);
        rp.shadowCastingMode = TreeShadowCasting[i];
        Graphics.RenderMesh(rp, TreeMesh, i, matrix);

    }
}

This function is called a million times in different scripts with different meshes. I want to find a way to make it less CPU-intensive. One current approach I am applying is to bring different objects under a common center point (multiple center points for different objects). Then, I calculate the distance between the center point and the camera position to execute a specific loop only for objects near the camera. Is there any alternative or more appropriate approach?

I have millions of objects that I want to render. I found that instantiate is not suitable, so I'm using Graphic.RenderMesh which is amazingly fast and light. But here comes the problem: it takes CPU resources because we are using a loop for each object and then rendering it:

void DrwaTree(Vector3 position, Quaternion rotation, Vector3 scale)
{
    Matrix4x4 matrix = Matrix4x4.TRS(position, rotation, scale);
      
    for (int i = 0; i < materials.Length; i++)
    {
        RenderParams rp = new RenderParams(materials[i]);
        rp.shadowCastingMode = TreeShadowCasting[i];
        Graphics.RenderMesh(rp, TreeMesh, i, matrix);
    }
}

This function is called a million times in different scripts with different meshes. I want to find a way to make it less CPU-intensive. 

One current approach I am applying is to bring different objects under a common center point (multiple center points for different objects). Then, I calculate the distance between the center point and the camera position to execute a specific loop only for objects near the camera. Is there any alternative or more appropriate approach?