tinymembench v0.4.9 (simple benchmark for memory throughput and latency)
==========================================================================
== Memory bandwidth tests ==
== ==
== Note 1: 1MB = 1000000 bytes ==
== Note 2: Results for 'copy' tests show how many bytes can be ==
== copied per second (adding together read and writen ==
== bytes would have provided twice higher numbers) ==
== Note 3: 2-pass copy means that we are using a small temporary buffer ==
== to first fetch data into it, and only then write it to the ==
== destination (source -> L1 cache, L1 cache -> destination) ==
== Note 4: If sample standard deviation exceeds 0.1%, it is shown in ==
== brackets ==
==========================================================================
C copy backwards : 12744.3 MB/s (0.7%)
C copy backwards (32 byte blocks) : 12732.3 MB/s (0.8%)
C copy backwards (64 byte blocks) : 12730.6 MB/s (0.6%)
C copy : 12749.2 MB/s (0.5%)
C copy prefetched (32 bytes step) : 12756.8 MB/s (0.5%)
C copy prefetched (64 bytes step) : 12768.9 MB/s (0.5%)
C 2-pass copy : 11952.6 MB/s (0.4%)
C 2-pass copy prefetched (32 bytes step) : 11523.8 MB/s (0.4%)
C 2-pass copy prefetched (64 bytes step) : 11625.2 MB/s (0.3%)
C fill : 18651.2 MB/s (0.3%)
C fill (shuffle within 16 byte blocks) : 18651.6 MB/s (0.3%)
C fill (shuffle within 32 byte blocks) : 18660.0 MB/s (0.3%)
C fill (shuffle within 64 byte blocks) : 18643.3 MB/s (0.2%)
---
standard memcpy : 12625.4 MB/s (0.6%)
standard memset : 18663.2 MB/s (0.3%)
---
NEON LDP/STP copy : 12593.5 MB/s (0.8%)
NEON LDP/STP copy pldl2strm (32 bytes step) : 12664.4 MB/s (0.5%)
NEON LDP/STP copy pldl2strm (64 bytes step) : 12620.4 MB/s (0.6%)
NEON LDP/STP copy pldl1keep (32 bytes step) : 12386.0 MB/s (0.8%)
NEON LDP/STP copy pldl1keep (64 bytes step) : 12260.9 MB/s (0.5%)
NEON LD1/ST1 copy : 12260.9 MB/s (0.4%)
NEON STP fill : 18639.6 MB/s (0.3%)
NEON STNP fill : 18650.1 MB/s (0.3%)
ARM LDP/STP copy : 12479.4 MB/s (0.6%)
ARM STP fill : 18622.6 MB/s (0.3%)
ARM STNP fill : 18624.6 MB/s (0.4%)
==========================================================================
== Memory latency test ==
== ==
== Average time is measured for random memory accesses in the buffers ==
== of different sizes. The larger is the buffer, the more significant ==
== are relative contributions of TLB, L1/L2 cache misses and SDRAM ==
== accesses. For extremely large buffer sizes we are expecting to see ==
== page table walk with several requests to SDRAM for almost every ==
== memory access (though 64MiB is not nearly large enough to experience ==
== this effect to its fullest). ==
== ==
== Note 1: All the numbers are representing extra time, which needs to ==
== be added to L1 cache latency. The cycle timings for L1 cache ==
== latency can be usually found in the processor documentation. ==
== Note 2: Dual random read means that we are simultaneously performing ==
== two independent memory accesses at a time. In the case if ==
== the memory subsystem can't handle multiple outstanding ==
== requests, dual random read has the same timings as two ==
== single reads performed one after another. ==
==========================================================================
block size : single random read / dual random read, [MADV_NOHUGEPAGE]
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.0 ns / 0.0 ns
65536 : 0.0 ns / 0.0 ns
131072 : 0.9 ns / 1.2 ns
262144 : 3.3 ns / 4.7 ns
524288 : 9.2 ns / 12.7 ns
1048576 : 12.8 ns / 15.6 ns
2097152 : 24.9 ns / 35.2 ns
4194304 : 74.2 ns / 104.0 ns
8388608 : 106.1 ns / 131.5 ns
16777216 : 128.2 ns / 144.8 ns
33554432 : 141.8 ns / 152.8 ns
67108864 : 155.8 ns / 163.1 ns
block size : single random read / dual random read, [MADV_HUGEPAGE]
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.0 ns / 0.0 ns
65536 : 0.0 ns / 0.0 ns
131072 : 0.9 ns / 1.2 ns
262144 : 2.7 ns / 3.6 ns
524288 : 8.0 ns / 11.5 ns
1048576 : 12.8 ns / 15.6 ns
2097152 : 20.5 ns / 26.8 ns
4194304 : 73.2 ns / 103.3 ns
8388608 : 106.0 ns / 130.6 ns
16777216 : 125.7 ns / 142.8 ns
33554432 : 136.2 ns / 148.0 ns
67108864 : 144.1 ns / 152.1 ns
Mongoose? What is the context of this test?