Guest User

OPi Zero Plus with more mature Armbian

a guest
Nov 17th, 2017
226
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
  1. root@orangepizeroplus:~/tinymembench# ./tinymembench
  2. tinymembench v0.4.9 (simple benchmark for memory throughput and latency)
  3.  
  4. ==========================================================================
  5. == Memory bandwidth tests ==
  6. == ==
  7. == Note 1: 1MB = 1000000 bytes ==
  8. == Note 2: Results for 'copy' tests show how many bytes can be ==
  9. == copied per second (adding together read and writen ==
  10. == bytes would have provided twice higher numbers) ==
  11. == Note 3: 2-pass copy means that we are using a small temporary buffer ==
  12. == to first fetch data into it, and only then write it to the ==
  13. == destination (source -> L1 cache, L1 cache -> destination) ==
  14. == Note 4: If sample standard deviation exceeds 0.1%, it is shown in ==
  15. == brackets ==
  16. ==========================================================================
  17.  
  18. C copy backwards : 857.0 MB/s (1.7%)
  19. C copy backwards (32 byte blocks) : 852.2 MB/s (1.2%)
  20. C copy backwards (64 byte blocks) : 859.7 MB/s (0.9%)
  21. C copy : 854.5 MB/s (1.3%)
  22. C copy prefetched (32 bytes step) : 683.1 MB/s
  23. C copy prefetched (64 bytes step) : 763.3 MB/s
  24. C 2-pass copy : 864.5 MB/s (0.6%)
  25. C 2-pass copy prefetched (32 bytes step) : 638.3 MB/s
  26. C 2-pass copy prefetched (64 bytes step) : 351.8 MB/s (0.2%)
  27. C fill : 2252.3 MB/s
  28. C fill (shuffle within 16 byte blocks) : 2252.1 MB/s (0.4%)
  29. C fill (shuffle within 32 byte blocks) : 2252.4 MB/s (1.0%)
  30. C fill (shuffle within 64 byte blocks) : 2253.0 MB/s (0.4%)
  31. ---
  32. standard memcpy : 875.1 MB/s (0.6%)
  33. standard memset : 2252.7 MB/s
  34. ---
  35. NEON LDP/STP copy : 865.7 MB/s (0.6%)
  36. NEON LDP/STP copy pldl2strm (32 bytes step) : 653.8 MB/s (0.6%)
  37. NEON LDP/STP copy pldl2strm (64 bytes step) : 774.2 MB/s
  38. NEON LDP/STP copy pldl1keep (32 bytes step) : 908.9 MB/s
  39. NEON LDP/STP copy pldl1keep (64 bytes step) : 909.2 MB/s
  40. NEON LD1/ST1 copy : 862.7 MB/s (1.3%)
  41. NEON STP fill : 2252.5 MB/s (0.6%)
  42. NEON STNP fill : 1881.2 MB/s (3.1%)
  43. ARM LDP/STP copy : 867.0 MB/s (0.4%)
  44. ARM STP fill : 2251.8 MB/s
  45. ARM STNP fill : 1876.4 MB/s (1.5%)
  46.  
  47. ==========================================================================
  48. == Memory latency test ==
  49. == ==
  50. == Average time is measured for random memory accesses in the buffers ==
  51. == of different sizes. The larger is the buffer, the more significant ==
  52. == are relative contributions of TLB, L1/L2 cache misses and SDRAM ==
  53. == accesses. For extremely large buffer sizes we are expecting to see ==
  54. == page table walk with several requests to SDRAM for almost every ==
  55. == memory access (though 64MiB is not nearly large enough to experience ==
  56. == this effect to its fullest). ==
  57. == ==
  58. == Note 1: All the numbers are representing extra time, which needs to ==
  59. == be added to L1 cache latency. The cycle timings for L1 cache ==
  60. == latency can be usually found in the processor documentation. ==
  61. == Note 2: Dual random read means that we are simultaneously performing ==
  62. == two independent memory accesses at a time. In the case if ==
  63. == the memory subsystem can't handle multiple outstanding ==
  64. == requests, dual random read has the same timings as two ==
  65. == single reads performed one after another. ==
  66. ==========================================================================
  67.  
  68. block size : single random read / dual random read, [MADV_NOHUGEPAGE]
  69. 1024 : 0.0 ns / 0.0 ns
  70. 2048 : 0.0 ns / 0.0 ns
  71. 4096 : 0.0 ns / 0.0 ns
  72. 8192 : 0.0 ns / 0.0 ns
  73. 16384 : 0.0 ns / 0.0 ns
  74. 32768 : 0.0 ns / 0.0 ns
  75. 65536 : 6.8 ns / 11.5 ns
  76. 131072 : 10.4 ns / 16.0 ns
  77. 262144 : 12.3 ns / 17.7 ns
  78. 524288 : 14.6 ns / 20.4 ns
  79. 1048576 : 101.0 ns / 156.7 ns
  80. 2097152 : 147.5 ns / 203.6 ns
  81. 4194304 : 178.4 ns / 229.0 ns
  82. 8388608 : 194.8 ns / 240.7 ns
  83. 16777216 : 205.0 ns / 248.2 ns
  84. 33554432 : 211.0 ns / 252.8 ns
  85. 67108864 : 214.7 ns / 256.1 ns
  86.  
  87. block size : single random read / dual random read, [MADV_HUGEPAGE]
  88. 1024 : 0.0 ns / 0.0 ns
  89. 2048 : 0.0 ns / 0.0 ns
  90. 4096 : 0.0 ns / 0.0 ns
  91. 8192 : 0.0 ns / 0.0 ns
  92. 16384 : 0.0 ns / 0.0 ns
  93. 32768 : 0.0 ns / 0.0 ns
  94. 65536 : 6.8 ns / 11.6 ns
  95. 131072 : 10.4 ns / 15.9 ns
  96. 262144 : 12.3 ns / 17.7 ns
  97. 524288 : 14.6 ns / 20.3 ns
  98. 1048576 : 101.0 ns / 156.8 ns
  99. 2097152 : 147.5 ns / 203.7 ns
  100. 4194304 : 178.4 ns / 229.1 ns
  101. 8388608 : 194.8 ns / 240.7 ns
  102. 16777216 : 204.8 ns / 247.9 ns
  103. 33554432 : 211.0 ns / 252.8 ns
  104. 67108864 : 214.8 ns / 256.3 ns
  105.  
  106.  
  107. root@orangepizeroplus:~# 7zr b
  108.  
  109. 7-Zip (a) [64] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21
  110. p7zip Version 16.02 (locale=en_US.UTF-8,Utf16=on,HugeFiles=on,64 bits,4 CPUs LE)
  111.  
  112. LE
  113. CPU Freq: 414 1006 1006 1005 1006 1006 1006 1005
  114.  
  115. RAM size: 482 MB, # CPU hardware threads: 4
  116. RAM usage: 450 MB, # Benchmark threads: 4
  117.  
  118. Compressing | Decompressing
  119. Dict Speed Usage R/U Rating | Speed Usage R/U Rating
  120. KiB/s % MIPS MIPS | KiB/s % MIPS MIPS
  121.  
  122. 22: 1557 306 495 1515 | 44917 399 961 3832
  123. 23: 1606 325 503 1637 | 43543 396 952 3768
  124. 24: 1305 307 457 1404 | 40686 389 919 3572
  125. ---------------------------------- | ------------------------------
  126. Avr: 313 485 1519 | 394 944 3724
  127. Tot: 354 714 2621
  128.  
  129.  
  130. root@orangepizeroplus:~# /bin/bash /tmp/sysbench.sh
  131. 408: execution time (avg/stddev): 89.2554/0.00
  132. 648: execution time (avg/stddev): 56.1463/0.00
  133. 816: execution time (avg/stddev): 44.3941/0.00
  134. 912: execution time (avg/stddev): 39.7057/0.00
  135. 960: execution time (avg/stddev): 37.7316/0.00
  136. 1008: execution time (avg/stddev): 35.9167/0.00
  137. 408: execution time (avg/stddev): 22.3622/0.00
  138. 648: execution time (avg/stddev): 14.1378/0.00
  139. 816: execution time (avg/stddev): 11.1106/0.00
  140. 912: execution time (avg/stddev): 9.9382/0.00
  141. 960: execution time (avg/stddev): 9.4391/0.00
  142. 1008: execution time (avg/stddev): 8.9818/0.00
  143.  
  144.  
  145. for i in 128 192 256 ; do openssl speed -elapsed -evp aes-${i}-cbc ; done
  146. type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes
  147. aes-128-cbc 69173.49k 218179.56k 457685.25k 655786.67k 749043.71k
  148. aes-192-cbc 66168.97k 196154.67k 379023.02k 509089.11k 565190.66k
  149. aes-256-cbc 64564.66k 183042.71k 332124.33k 427910.83k 466774.70k
RAW Paste Data