Page 1 of 1

SSE 4-Ray vs. Sphere?

Posted: Thu Aug 08, 2013 8:48 pm
I'm finally getting with the program and working on learning SSE, but I'm having trouble trying to build a 4 ray vs. 1 sphere intersection function. While it's trivial in single-ray land, when working with SIMD instructions I'm getting pretty confused about how one goes about handling branching.

Does anybody have any pointers to where I can read more about the topic, or (better yet) have any example code that I could dig into?

Thanks!

Re: SSE 4-Ray vs. Sphere?

Posted: Thu Aug 08, 2013 8:52 pm
This appears to show how to do this:

http://felix.abecassis.me/2012/08/sse-v ... onal-code/

Re: SSE 4-Ray vs. Sphere?

Posted: Fri Aug 09, 2013 7:33 pm
Thanks a ton, that was exactly the nudge I needed.

For example, here's a 4-ray vs. sphere:

Code: Select all

`static v4sf intersect(const ray4_t *ray) {  /* Compute discriminant. */  const v4sf b = -dot(&(ray->origin), &(ray->direction)),             d = b * b - sq(&(ray->origin)) + V4SF_ONE,             t = b - __builtin_ia32_sqrtps(d);  /* Which rays hit the sphere? (Note that this works since any rays that   * didn't hit the sphere will be NaN, because we had to sqrt a negative   * number. Comparing that to zero will be false as a result.) */  const v4si mask = (v4si) __builtin_ia32_cmpgeps(t, V4SF_ZERO);  /* Return the rays that hit the sphere, or infinity otherwise. */  return __builtin_ia32_orps(    __builtin_ia32_andps((v4sf) mask, t),    __builtin_ia32_andnps((v4sf) mask, V4SF_INFINITY)  );}`

(Yeah, yeah, the sphere is always a unit sphere centered on the origin and we only check against the outside, but one thing at a time. I also don't know if it's faster to bail early if all of the discriminants are negative or not. I'll have to test.)