API tools faq
paste
Guest User

Performance results dot product via C++ algorithms, HPX

a guest
Jul 11th, 2018
134
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 10.95 KB | None | 0 0
raw download clone embed print report
  1. number = 1024.
  2. 8.00 kB
  3. dot(x1, x2, number) : 3.308 cycles per operation (best) 2.419 bytes per cycle (best) 3.322 cycles per operation (avg)
  4. dot128(x1, x2, number) : 0.843 cycles per operation (best) 9.492 bytes per cycle (best) 0.852 cycles per operation (avg)
  5. dot128dt(x1, x2, number) : 1.919 cycles per operation (best) 4.169 bytes per cycle (best) 1.985 cycles per operation (avg)
  6. dot256(x1, x2, number) : 0.520 cycles per operation (best) 15.398 bytes per cycle (best) 0.553 cycles per operation (avg)
  7. dot128fma(x1, x2, number) : 0.839 cycles per operation (best) 9.537 bytes per cycle (best) 0.854 cycles per operation (avg)
  8. dot256fma(x1, x2, number) : 0.526 cycles per operation (best) 15.199 bytes per cycle (best) 0.554 cycles per operation (avg)
  9. dot_inner_product(x1, x2, number) : 3.538 cycles per operation (best) 2.261 bytes per cycle (best) 3.576 cycles per operation (avg)
  10. dot_transform_reduce(x1, x2, number) : 94.146 cycles per operation (best) 0.085 bytes per cycle (best) 183.408 cycles per operation (avg)
  11.  
  12. number = 1024.
  13. 8.00 kB
  14. dot(x1, x2, number) : 3.313 cycles per operation (best) 2.414 bytes per cycle (best) 3.330 cycles per operation (avg)
  15. dot128(x1, x2, number) : 0.850 cycles per operation (best) 9.416 bytes per cycle (best) 0.862 cycles per operation (avg)
  16. dot128dt(x1, x2, number) : 1.871 cycles per operation (best) 4.276 bytes per cycle (best) 1.992 cycles per operation (avg)
  17. dot256(x1, x2, number) : 0.495 cycles per operation (best) 16.158 bytes per cycle (best) 0.558 cycles per operation (avg)
  18. dot128fma(x1, x2, number) : 0.849 cycles per operation (best) 9.427 bytes per cycle (best) 0.862 cycles per operation (avg)
  19. dot256fma(x1, x2, number) : 0.525 cycles per operation (best) 15.227 bytes per cycle (best) 0.563 cycles per operation (avg)
  20. dot_inner_product(x1, x2, number) : 3.534 cycles per operation (best) 2.264 bytes per cycle (best) 3.581 cycles per operation (avg)
  21. dot_transform_reduce(x1, x2, number) : 119.325 cycles per operation (best) 0.067 bytes per cycle (best) 199.220 cycles per operation (avg)
  22.  
  23. number = 2097152.
  24. 16.00 MB
  25. dot(x1, x2, number) : 3.472 cycles per operation (best) 2.304 bytes per cycle (best) 3.981 cycles per operation (avg)
  26. dot128(x1, x2, number) : 1.067 cycles per operation (best) 7.497 bytes per cycle (best) 1.651 cycles per operation (avg)
  27. dot128dt(x1, x2, number) : 1.472 cycles per operation (best) 5.434 bytes per cycle (best) 2.436 cycles per operation (avg)
  28. dot256(x1, x2, number) : 0.966 cycles per operation (best) 8.280 bytes per cycle (best) 1.525 cycles per operation (avg)
  29. dot128fma(x1, x2, number) : 1.035 cycles per operation (best) 7.728 bytes per cycle (best) 1.333 cycles per operation (avg)
  30. dot256fma(x1, x2, number) : 0.971 cycles per operation (best) 8.239 bytes per cycle (best) 1.042 cycles per operation (avg)
  31. dot_inner_product(x1, x2, number) : 3.491 cycles per operation (best) 2.291 bytes per cycle (best) 3.732 cycles per operation (avg)
  32. dot_transform_reduce(x1, x2, number) : 0.724 cycles per operation (best) 11.043 bytes per cycle (best) 1.325 cycles per operation (avg)
  33.  
  34. number = 4194304.
  35. 32.00 MB
  36. dot(x1, x2, number) : 3.437 cycles per operation (best) 2.328 bytes per cycle (best) 3.523 cycles per operation (avg)
  37. dot128(x1, x2, number) : 1.115 cycles per operation (best) 7.177 bytes per cycle (best) 1.187 cycles per operation (avg)
  38. dot128dt(x1, x2, number) : 1.455 cycles per operation (best) 5.497 bytes per cycle (best) 1.944 cycles per operation (avg)
  39. dot256(x1, x2, number) : 0.916 cycles per operation (best) 8.735 bytes per cycle (best) 1.043 cycles per operation (avg)
  40. dot128fma(x1, x2, number) : 0.997 cycles per operation (best) 8.024 bytes per cycle (best) 1.104 cycles per operation (avg)
  41. dot256fma(x1, x2, number) : 1.043 cycles per operation (best) 7.667 bytes per cycle (best) 1.075 cycles per operation (avg)
  42. dot_inner_product(x1, x2, number) : 3.263 cycles per operation (best) 2.452 bytes per cycle (best) 3.628 cycles per operation (avg)
  43. dot_transform_reduce(x1, x2, number) : 0.756 cycles per operation (best) 10.578 bytes per cycle (best) 1.045 cycles per operation (avg)
  44.  
  45. number = 8388608.
  46. 64.00 MB
  47. dot(x1, x2, number) : 3.437 cycles per operation (best) 2.328 bytes per cycle (best) 3.520 cycles per operation (avg)
  48. dot128(x1, x2, number) : 1.057 cycles per operation (best) 7.569 bytes per cycle (best) 1.141 cycles per operation (avg)
  49. dot128dt(x1, x2, number) : 1.481 cycles per operation (best) 5.401 bytes per cycle (best) 1.876 cycles per operation (avg)
  50. dot256(x1, x2, number) : 0.976 cycles per operation (best) 8.195 bytes per cycle (best) 1.071 cycles per operation (avg)
  51. dot128fma(x1, x2, number) : 1.054 cycles per operation (best) 7.592 bytes per cycle (best) 1.145 cycles per operation (avg)
  52. dot256fma(x1, x2, number) : 0.946 cycles per operation (best) 8.459 bytes per cycle (best) 1.065 cycles per operation (avg)
  53. dot_inner_product(x1, x2, number) : 3.364 cycles per operation (best) 2.378 bytes per cycle (best) 3.582 cycles per operation (avg)
  54. dot_transform_reduce(x1, x2, number) : 0.770 cycles per operation (best) 10.392 bytes per cycle (best) 1.246 cycles per operation (avg)
  55.  
  56. number = 16777216.
  57. 128.00 MB
  58. dot(x1, x2, number) : 3.208 cycles per operation (best) 2.494 bytes per cycle (best) 3.501 cycles per operation (avg)
  59. dot128(x1, x2, number) : 1.059 cycles per operation (best) 7.552 bytes per cycle (best) 1.154 cycles per operation (avg)
  60. dot128dt(x1, x2, number) : 1.453 cycles per operation (best) 5.505 bytes per cycle (best) 1.769 cycles per operation (avg)
  61. dot256(x1, x2, number) : 0.992 cycles per operation (best) 8.068 bytes per cycle (best) 1.077 cycles per operation (avg)
  62. dot128fma(x1, x2, number) : 1.047 cycles per operation (best) 7.642 bytes per cycle (best) 1.147 cycles per operation (avg)
  63. dot256fma(x1, x2, number) : 0.986 cycles per operation (best) 8.116 bytes per cycle (best) 1.071 cycles per operation (avg)
  64. dot_inner_product(x1, x2, number) : 3.204 cycles per operation (best) 2.497 bytes per cycle (best) 3.556 cycles per operation (avg)
  65. dot_transform_reduce(x1, x2, number) : 0.769 cycles per operation (best) 10.404 bytes per cycle (best) 0.868 cycles per operation (avg)
  66.  
  67. number = 33554432.
  68. 256.00 MB
  69. dot(x1, x2, number) : 3.453 cycles per operation (best) 2.317 bytes per cycle (best) 3.515 cycles per operation (avg)
  70. dot128(x1, x2, number) : 1.032 cycles per operation (best) 7.754 bytes per cycle (best) 1.134 cycles per operation (avg)
  71. dot128dt(x1, x2, number) : 1.427 cycles per operation (best) 5.608 bytes per cycle (best) 1.690 cycles per operation (avg)
  72. dot256(x1, x2, number) : 0.998 cycles per operation (best) 8.014 bytes per cycle (best) 1.060 cycles per operation (avg)
  73. dot128fma(x1, x2, number) : 1.044 cycles per operation (best) 7.664 bytes per cycle (best) 1.115 cycles per operation (avg)
  74. dot256fma(x1, x2, number) : 0.991 cycles per operation (best) 8.074 bytes per cycle (best) 1.050 cycles per operation (avg)
  75. dot_inner_product(x1, x2, number) : 3.255 cycles per operation (best) 2.457 bytes per cycle (best) 3.445 cycles per operation (avg)
  76. dot_transform_reduce(x1, x2, number) : 0.766 cycles per operation (best) 10.443 bytes per cycle (best) 0.796 cycles per operation (avg)
  77.  
  78. number = 134217728.
  79. 1024.00 MB
  80. dot(x1, x2, number) : 3.383 cycles per operation (best) 2.365 bytes per cycle (best) 3.478 cycles per operation (avg)
  81. dot128(x1, x2, number) : 1.034 cycles per operation (best) 7.737 bytes per cycle (best) 1.084 cycles per operation (avg)
  82. dot128dt(x1, x2, number) : 1.395 cycles per operation (best) 5.733 bytes per cycle (best) 1.555 cycles per operation (avg)
  83. dot256(x1, x2, number) : 0.962 cycles per operation (best) 8.314 bytes per cycle (best) 1.025 cycles per operation (avg)
  84. dot128fma(x1, x2, number) : 1.001 cycles per operation (best) 7.990 bytes per cycle (best) 1.063 cycles per operation (avg)
  85. dot256fma(x1, x2, number) : 0.990 cycles per operation (best) 8.078 bytes per cycle (best) 1.023 cycles per operation (avg)
  86. dot_inner_product(x1, x2, number) : 3.187 cycles per operation (best) 2.510 bytes per cycle (best) 3.337 cycles per operation (avg)
  87. dot_transform_reduce(x1, x2, number) : 0.758 cycles per operation (best) 10.558 bytes per cycle (best) 0.805 cycles per operation (avg)
Advertisement
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!