Raspberry Pi 2 (BCM2836) - ssvb/tinymembench GitHub Wiki
processor : 0 model name : ARMv7 Processor rev 5 (v7l) BogoMIPS : 57.60 Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm CPU implementer : 0x41 CPU architecture: 7 CPU variant : 0x0 CPU part : 0xc07 CPU revision : 5 processor : 1 model name : ARMv7 Processor rev 5 (v7l) BogoMIPS : 57.60 Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm CPU implementer : 0x41 CPU architecture: 7 CPU variant : 0x0 CPU part : 0xc07 CPU revision : 5 processor : 2 model name : ARMv7 Processor rev 5 (v7l) BogoMIPS : 57.60 Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm CPU implementer : 0x41 CPU architecture: 7 CPU variant : 0x0 CPU part : 0xc07 CPU revision : 5 processor : 3 model name : ARMv7 Processor rev 5 (v7l) BogoMIPS : 57.60 Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm CPU implementer : 0x41 CPU architecture: 7 CPU variant : 0x0 CPU part : 0xc07 CPU revision : 5 Hardware : BCM2709 Revision : a01041 Serial : 00000000ec797b2c
The standard 900MHz configuration without any overclocking or config.txt tweaks (2015-01-31-raspbian) and no monitor connected
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 : 819.5 MB/s
C copy : 970.1 MB/s (5.4%)
C copy prefetched (32 bytes step) : 920.6 MB/s
C copy prefetched (64 bytes step) : 920.2 MB/s
C 2-pass copy : 612.6 MB/s (0.3%)
C 2-pass copy prefetched (32 bytes step) : 624.8 MB/s
C 2-pass copy prefetched (64 bytes step) : 624.8 MB/s (0.8%)
C fill : 1209.9 MB/s (0.5%)
---
standard memcpy : 1006.9 MB/s
standard memset : 1210.3 MB/s (0.5%)
---
NEON read : 1341.3 MB/s (5.3%)
NEON read prefetched (32 bytes step) : 1370.8 MB/s
NEON read prefetched (64 bytes step) : 1364.3 MB/s (5.6%)
NEON read 2 data streams : 386.0 MB/s (0.1%)
NEON read 2 data streams prefetched (32 bytes step) : 718.6 MB/s
NEON read 2 data streams prefetched (64 bytes step) : 757.6 MB/s
NEON copy : 990.0 MB/s
NEON copy prefetched (32 bytes step) : 954.3 MB/s
NEON copy prefetched (64 bytes step) : 1028.2 MB/s
NEON unrolled copy : 941.5 MB/s
NEON unrolled copy prefetched (32 bytes step) : 916.5 MB/s
NEON unrolled copy prefetched (64 bytes step) : 983.7 MB/s
NEON copy backwards : 825.9 MB/s
NEON copy backwards prefetched (32 bytes step) : 784.6 MB/s
NEON copy backwards prefetched (64 bytes step) : 861.2 MB/s
NEON 2-pass copy : 628.8 MB/s
NEON 2-pass copy prefetched (32 bytes step) : 646.2 MB/s
NEON 2-pass copy prefetched (64 bytes step) : 652.2 MB/s
NEON unrolled 2-pass copy : 589.8 MB/s
NEON unrolled 2-pass copy prefetched (32 bytes step) : 583.8 MB/s
NEON unrolled 2-pass copy prefetched (64 bytes step) : 613.9 MB/s
NEON fill : 1210.4 MB/s
NEON fill backwards : 1209.8 MB/s
VFP copy : 950.9 MB/s
VFP 2-pass copy : 590.1 MB/s
ARM fill (STRD) : 1209.2 MB/s
ARM fill (STM with 8 registers) : 1210.3 MB/s
ARM fill (STM with 4 registers) : 1210.4 MB/s
ARM copy prefetched (incr pld) : 965.5 MB/s
ARM copy prefetched (wrap pld) : 833.2 MB/s
ARM 2-pass copy prefetched (incr pld) : 638.9 MB/s
ARM 2-pass copy prefetched (wrap pld) : 609.4 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 : 0.0 ns / 0.1 ns
65536 : 6.5 ns / 11.6 ns
131072 : 10.0 ns / 16.8 ns
262144 : 11.8 ns / 19.1 ns
524288 : 16.7 ns / 26.8 ns
1048576 : 89.2 ns / 142.3 ns
2097152 : 134.8 ns / 190.1 ns
4194304 : 158.4 ns / 208.7 ns
8388608 : 172.0 ns / 218.1 ns
16777216 : 183.0 ns / 229.2 ns
33554432 : 193.0 ns / 243.0 ns
67108864 : 209.1 ns / 272.4 ns
The standard 900MHz configuration without any overclocking or config.txt tweaks (2015-01-31-raspbian) and a 1920x1080-32@60Hz HDMI monitor connected
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 : 774.0 MB/s (1.6%)
C copy : 898.5 MB/s
C copy prefetched (32 bytes step) : 862.4 MB/s (7.1%)
C copy prefetched (64 bytes step) : 862.3 MB/s
C 2-pass copy : 568.8 MB/s (0.4%)
C 2-pass copy prefetched (32 bytes step) : 584.8 MB/s
C 2-pass copy prefetched (64 bytes step) : 584.5 MB/s (0.4%)
C fill : 1134.5 MB/s (0.2%)
---
standard memcpy : 922.8 MB/s (1.2%)
standard memset : 1136.1 MB/s
---
NEON read : 1245.1 MB/s (0.7%)
NEON read prefetched (32 bytes step) : 1281.0 MB/s
NEON read prefetched (64 bytes step) : 1283.5 MB/s (0.6%)
NEON read 2 data streams : 369.5 MB/s (0.1%)
NEON read 2 data streams prefetched (32 bytes step) : 688.4 MB/s
NEON read 2 data streams prefetched (64 bytes step) : 723.8 MB/s
NEON copy : 910.5 MB/s
NEON copy prefetched (32 bytes step) : 874.4 MB/s
NEON copy prefetched (64 bytes step) : 946.8 MB/s
NEON unrolled copy : 867.0 MB/s
NEON unrolled copy prefetched (32 bytes step) : 861.7 MB/s
NEON unrolled copy prefetched (64 bytes step) : 916.7 MB/s
NEON copy backwards : 742.3 MB/s
NEON copy backwards prefetched (32 bytes step) : 723.4 MB/s
NEON copy backwards prefetched (64 bytes step) : 762.7 MB/s
NEON 2-pass copy : 583.0 MB/s
NEON 2-pass copy prefetched (32 bytes step) : 605.6 MB/s
NEON 2-pass copy prefetched (64 bytes step) : 612.5 MB/s
NEON unrolled 2-pass copy : 547.6 MB/s
NEON unrolled 2-pass copy prefetched (32 bytes step) : 543.9 MB/s
NEON unrolled 2-pass copy prefetched (64 bytes step) : 572.0 MB/s
NEON fill : 1136.4 MB/s
NEON fill backwards : 1136.7 MB/s
VFP copy : 873.9 MB/s
VFP 2-pass copy : 548.5 MB/s
ARM fill (STRD) : 1133.1 MB/s
ARM fill (STM with 8 registers) : 1136.6 MB/s
ARM fill (STM with 4 registers) : 1135.8 MB/s
ARM copy prefetched (incr pld) : 894.0 MB/s
ARM copy prefetched (wrap pld) : 736.6 MB/s
ARM 2-pass copy prefetched (incr pld) : 594.5 MB/s
ARM 2-pass copy prefetched (wrap pld) : 559.6 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 : 0.2 ns / 0.2 ns
65536 : 6.5 ns / 11.7 ns
131072 : 10.0 ns / 16.9 ns
262144 : 11.8 ns / 19.2 ns
524288 : 16.4 ns / 26.0 ns
1048576 : 95.3 ns / 151.8 ns
2097152 : 143.3 ns / 201.9 ns
4194304 : 168.2 ns / 221.5 ns
8388608 : 182.9 ns / 232.1 ns
16777216 : 194.1 ns / 243.1 ns
33554432 : 205.3 ns / 259.0 ns
67108864 : 225.5 ns / 296.8 ns