gs702c - ssvb/tinymembench GitHub Wiki
ATM7029B?
tinymembench v0.3.9 (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 : 260.6 MB/s (0.4%)
C copy : 323.8 MB/s (0.7%)
C copy prefetched (32 bytes step) : 372.8 MB/s
C copy prefetched (64 bytes step) : 374.4 MB/s (0.5%)
C 2-pass copy : 292.0 MB/s (0.3%)
C 2-pass copy prefetched (32 bytes step) : 325.6 MB/s
C 2-pass copy prefetched (64 bytes step) : 325.5 MB/s
C fill : 1445.6 MB/s (1.9%)
---
standard memcpy : 332.5 MB/s (0.5%)
standard memset : 1434.4 MB/s
---
NEON read : 621.2 MB/s
NEON read prefetched (32 bytes step) : 950.8 MB/s
NEON read prefetched (64 bytes step) : 1069.2 MB/s
NEON read 2 data streams : 761.7 MB/s
NEON read 2 data streams prefetched (32 bytes step) : 1253.5 MB/s (0.2%)
NEON read 2 data streams prefetched (64 bytes step) : 1020.1 MB/s
NEON copy : 322.1 MB/s
NEON copy prefetched (32 bytes step) : 392.4 MB/s (1.0%)
NEON copy prefetched (64 bytes step) : 336.3 MB/s (0.1%)
NEON unrolled copy : 329.7 MB/s
NEON unrolled copy prefetched (32 bytes step) : 364.4 MB/s
NEON unrolled copy prefetched (64 bytes step) : 346.2 MB/s
NEON copy backwards : 280.7 MB/s (0.1%)
NEON copy backwards prefetched (32 bytes step) : 380.8 MB/s (0.7%)
NEON copy backwards prefetched (64 bytes step) : 321.4 MB/s
NEON 2-pass copy : 298.4 MB/s
NEON 2-pass copy prefetched (32 bytes step) : 353.6 MB/s
NEON 2-pass copy prefetched (64 bytes step) : 324.5 MB/s
NEON unrolled 2-pass copy : 293.4 MB/s
NEON unrolled 2-pass copy prefetched (32 bytes step) : 333.5 MB/s (0.2%)
NEON unrolled 2-pass copy prefetched (64 bytes step) : 322.4 MB/s (0.2%)
NEON fill : 1435.4 MB/s
NEON fill backwards : 1434.1 MB/s
VFP copy : 327.8 MB/s
VFP 2-pass copy : 297.6 MB/s
ARM fill (STRD) : 1435.5 MB/s
ARM fill (STM with 8 registers) : 1433.0 MB/s
ARM fill (STM with 4 registers) : 1435.2 MB/s
ARM copy prefetched (incr pld) : 378.2 MB/s
ARM copy prefetched (wrap pld) : 335.5 MB/s
ARM 2-pass copy prefetched (incr pld) : 343.4 MB/s
ARM 2-pass copy prefetched (wrap pld) : 311.7 MB/s
==========================================================================
== 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
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 : 11.9 ns / 19.1 ns
65536 : 18.6 ns / 26.1 ns
131072 : 22.2 ns / 28.8 ns
262144 : 24.0 ns / 30.0 ns
524288 : 28.5 ns / 35.8 ns
1048576 : 89.8 ns / 147.0 ns
2097152 : 121.4 ns / 197.2 ns
4194304 : 139.9 ns / 224.1 ns
8388608 : 154.2 ns / 243.1 ns
16777216 : 165.9 ns / 259.2 ns
33554432 : 176.2 ns / 276.5 ns
67108864 : 189.5 ns / 303.5 ns