Pinned Loading

1. kern.c kern.c

   1

   #define EXECUTE(j)                                                        \

   2

       {                                                                     \

   3

           __m256 floats1 = _mm256_load_ps(&datar[b + j]);                   \

   4

           __m256 floats2 = _mm256_load_ps(&datac[b + j]);                   \

   5

           res1 = _mm256_add_ps(res1, _mm256_mul_ps(revKernel[j], floats1)); \

2. partial_results.txt partial_results.txt

   1

   8 80000 63 81 123 61

   2

   16 80000 122 139 249 106

   3

   24 80000 120 312 478 794

   4

   32 80000 151 348 584 1085

   5

   64 80000 331 574 1111 2518