The raytracing acceleration structure used is a bounding volume hierarchy. The original code is based on an implementation from Nvidia (BHV2, BHV4) and Embree (Hair BVH builder – oriented SAH, QBVH traversal, BVH nodes intersection, Triangles intersection).
We added this settings to Blender:
I used the scenes from this page: https://code.blender.org/2016/02/new-cycles-benchmark/
- 2x Intel® Xeon™ E5-2680v3 (HASWELLS)
- Marconi-A2 CINECA
- Intel® Xeon Phi™ 7250 (KNL)
Blender 2.79 was compiled with GCC. You cand find more tests on Intel Xeon Phi here.