sbc-bench v0.7.5 LeMaker Banana Pi (Sun, 06 Dec 2020 06:24:37 +0000) Distributor ID: Debian Description: Debian GNU/Linux 10 (buster) Release: 10 Codename: buster Armbian release info: BOARD=bananapi BOARD_NAME="Banana Pi" BOARDFAMILY=sun7i BUILD_REPOSITORY_URL=https://github.com/armbian/build BUILD_REPOSITORY_COMMIT=b0760915-dirty DISTRIBUTION_CODENAME=buster DISTRIBUTION_STATUS=supported VERSION=20.11 LINUXFAMILY=sunxi BRANCH=current ARCH=arm IMAGE_TYPE=stable BOARD_TYPE=conf INITRD_ARCH=arm KERNEL_IMAGE_TYPE=Image /usr/bin/gcc (Debian 8.3.0-6) 8.3.0 Uptime: 06:24:37 up 47 min, 2 users, load average: 1.46, 1.71, 1.28 Linux 5.8.16-sunxi (bananapi) 12/06/2020 _armv7l_ (2 CPU) avg-cpu: %user %nice %system %iowait %steal %idle 12.46 0.00 7.83 2.14 0.00 77.57 Device tps kB_read/s kB_wrtn/s kB_read kB_wrtn mmcblk0 2.41 78.76 56.41 224068 160468 zram0 0.13 0.26 0.75 752 2140 zram1 0.10 0.42 0.00 1184 4 total used free shared buff/cache available Mem: 967Mi 82Mi 610Mi 2.0Mi 274Mi 851Mi Swap: 483Mi 0B 483Mi Filename Type Size Used Priority /dev/zram1 partition 495108 0 5 ########################################################################## Checking cpufreq OPP: Cpufreq OPP: 1200 Measured: 1197.369/1198.355/1196.232 Cpufreq OPP: 1152 Measured: 1148.513/1013.560/1149.766 Cpufreq OPP: 1104 Measured: 1101.253/1102.463/1101.711 Cpufreq OPP: 1056 Measured: 1052.086/1171.717/1053.105 Cpufreq OPP: 1008 Measured: 1005.027/1006.643/1004.954 Cpufreq OPP: 960 Measured: 956.277/957.252/957.529 Cpufreq OPP: 912 Measured: 909.598/903.897/909.948 Cpufreq OPP: 864 Measured: 861.648/861.882/861.280 Cpufreq OPP: 720 Measured: 717.585/713.789/717.110 Cpufreq OPP: 528 Measured: 525.477/525.719/524.832 Cpufreq OPP: 312 Measured: 309.278/310.160/306.964 Cpufreq OPP: 144 Measured: 141.712/140.963/141.761 ########################################################################## 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 : 213.9 MB/s (0.8%) C copy backwards (32 byte blocks) : 544.6 MB/s (10.0%) C copy backwards (64 byte blocks) : 551.7 MB/s (7.8%) C copy : 518.2 MB/s (3.5%) C copy prefetched (32 bytes step) : 547.4 MB/s (3.9%) C copy prefetched (64 bytes step) : 550.7 MB/s (3.2%) C 2-pass copy : 523.0 MB/s C 2-pass copy prefetched (32 bytes step) : 525.6 MB/s (3.3%) C 2-pass copy prefetched (64 bytes step) : 526.2 MB/s (3.2%) C fill : 1416.8 MB/s (0.7%) C fill (shuffle within 16 byte blocks) : 1417.4 MB/s (0.6%) C fill (shuffle within 32 byte blocks) : 279.7 MB/s (0.5%) C fill (shuffle within 64 byte blocks) : 292.0 MB/s (0.6%) --- standard memcpy : 369.6 MB/s (1.7%) standard memset : 1416.7 MB/s (4.2%) --- NEON read : 862.6 MB/s (5.1%) NEON read prefetched (32 bytes step) : 986.0 MB/s (1.9%) NEON read prefetched (64 bytes step) : 1015.5 MB/s (5.4%) NEON read 2 data streams : 305.5 MB/s (0.3%) NEON read 2 data streams prefetched (32 bytes step) : 571.6 MB/s (0.4%) NEON read 2 data streams prefetched (64 bytes step) : 586.5 MB/s (0.6%) NEON copy : 518.5 MB/s (0.4%) NEON copy prefetched (32 bytes step) : 526.8 MB/s (6.0%) NEON copy prefetched (64 bytes step) : 557.0 MB/s (0.6%) NEON unrolled copy : 526.4 MB/s (9.2%) NEON unrolled copy prefetched (32 bytes step) : 541.2 MB/s (10.5%) NEON unrolled copy prefetched (64 bytes step) : 551.0 MB/s (9.6%) NEON copy backwards : 544.6 MB/s (9.8%) NEON copy backwards prefetched (32 bytes step) : 578.7 MB/s (7.8%) NEON copy backwards prefetched (64 bytes step) : 637.8 MB/s (0.8%) NEON 2-pass copy : 523.6 MB/s NEON 2-pass copy prefetched (32 bytes step) : 526.2 MB/s (3.2%) NEON 2-pass copy prefetched (64 bytes step) : 527.2 MB/s (0.4%) NEON unrolled 2-pass copy : 522.3 MB/s (0.9%) NEON unrolled 2-pass copy prefetched (32 bytes step) : 521.3 MB/s (3.3%) NEON unrolled 2-pass copy prefetched (64 bytes step) : 523.5 MB/s (0.3%) NEON fill : 1417.1 MB/s (0.6%) NEON fill backwards : 1567.1 MB/s (1.3%) VFP copy : 526.6 MB/s VFP 2-pass copy : 521.2 MB/s ARM fill (STRD) : 1416.7 MB/s (0.7%) ARM fill (STM with 8 registers) : 1417.5 MB/s (0.7%) ARM fill (STM with 4 registers) : 1417.0 MB/s (4.2%) ARM copy prefetched (incr pld) : 559.0 MB/s (6.3%) ARM copy prefetched (wrap pld) : 522.8 MB/s (8.0%) ARM 2-pass copy prefetched (incr pld) : 526.1 MB/s (0.5%) ARM 2-pass copy prefetched (wrap pld) : 523.9 MB/s (3.2%) ========================================================================== == Framebuffer read tests. == == == == Many ARM devices use a part of the system memory as the framebuffer, == == typically mapped as uncached but with write-combining enabled. == == Writes to such framebuffers are quite fast, but reads are much == == slower and very sensitive to the alignment and the selection of == == CPU instructions which are used for accessing memory. == == == == Many x86 systems allocate the framebuffer in the GPU memory, == == accessible for the CPU via a relatively slow PCI-E bus. Moreover, == == PCI-E is asymmetric and handles reads a lot worse than writes. == == == == If uncached framebuffer reads are reasonably fast (at least 100 MB/s == == or preferably >300 MB/s), then using the shadow framebuffer layer == == is not necessary in Xorg DDX drivers, resulting in a nice overall == == performance improvement. For example, the xf86-video-fbturbo DDX == == uses this trick. == ========================================================================== NEON read (from framebuffer) : 44.3 MB/s (0.3%) NEON copy (from framebuffer) : 43.1 MB/s (0.4%) NEON 2-pass copy (from framebuffer) : 42.7 MB/s NEON unrolled copy (from framebuffer) : 42.5 MB/s (0.3%) NEON 2-pass unrolled copy (from framebuffer) : 42.5 MB/s (0.3%) VFP copy (from framebuffer) : 230.4 MB/s (0.2%) VFP 2-pass copy (from framebuffer) : 242.1 MB/s (0.3%) ARM copy (from framebuffer) : 154.6 MB/s (0.3%) ARM 2-pass copy (from framebuffer) : 146.6 MB/s (0.2%) ========================================================================== == 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 / 0.0 ns 65536 : 5.3 ns / 9.1 ns 131072 : 8.1 ns / 12.7 ns 262144 : 13.2 ns / 19.5 ns 524288 : 119.1 ns / 190.0 ns 1048576 : 178.2 ns / 250.5 ns 2097152 : 213.4 ns / 281.8 ns 4194304 : 231.8 ns / 290.8 ns 8388608 : 243.7 ns / 300.8 ns 16777216 : 255.9 ns / 316.5 ns 33554432 : 274.0 ns / 352.6 ns 67108864 : 335.8 ns / 459.1 ns ########################################################################## OpenSSL 1.1.1d, built on 10 Sep 2019 type 16 bytes 64 bytes 256 bytes 1024 bytes 8192 bytes 16384 bytes aes-128-cbc 17970.23k 24768.66k 27746.65k 28490.41k 29906.26k 28857.69k aes-128-cbc 17425.51k 25558.34k 27746.65k 28607.15k 29783.38k 28819.46k aes-192-cbc 15685.29k 21509.74k 24538.71k 24408.41k 24600.58k 25389.74k aes-192-cbc 15689.94k 21509.01k 23767.64k 25199.27k 24603.31k 24592.38k aes-256-cbc 15511.44k 20494.61k 22599.59k 23138.99k 23232.51k 23303.51k aes-256-cbc 15518.34k 20670.51k 22391.13k 23136.60k 23298.05k 22975.70k ########################################################################## 7-Zip (a) [32] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21 p7zip Version 16.02 (locale=en_US.UTF-8,Utf16=on,HugeFiles=on,32 bits,2 CPUs LE) LE CPU Freq: 1123 1186 1187 1188 1188 1187 1188 1179 1173 RAM size: 967 MB, # CPU hardware threads: 2 RAM usage: 441 MB, # Benchmark threads: 2 Compressing | Decompressing Dict Speed Usage R/U Rating | Speed Usage R/U Rating KiB/s % MIPS MIPS | KiB/s % MIPS MIPS 22: 461 100 449 449 | 11129 100 950 950 23: 447 100 456 456 | 10868 100 941 941 24: 432 100 465 465 | 10580 100 929 929 25: 417 100 477 477 | 10064 100 896 896 ---------------------------------- | ------------------------------ Avr: 100 462 462 | 100 929 929 Tot: 100 696 695 ########################################################################## 7-Zip (a) [32] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21 p7zip Version 16.02 (locale=en_US.UTF-8,Utf16=on,HugeFiles=on,32 bits,2 CPUs LE) LE CPU Freq: 1129 1148 1182 1184 1184 1150 1184 1181 1114 RAM size: 967 MB, # CPU hardware threads: 2 RAM usage: 441 MB, # Benchmark threads: 2 Compressing | Decompressing Dict Speed Usage R/U Rating | Speed Usage R/U Rating KiB/s % MIPS MIPS | KiB/s % MIPS MIPS 22: 694 160 423 675 | 20952 189 944 1789 23: 671 160 427 684 | 19279 178 936 1669 24: 647 161 433 696 | 19126 182 921 1679 25: 645 166 443 737 | 18438 186 884 1641 ---------------------------------- | ------------------------------ Avr: 162 432 698 | 184 921 1695 Tot: 173 676 1196 7-Zip (a) [32] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21 p7zip Version 16.02 (locale=en_US.UTF-8,Utf16=on,HugeFiles=on,32 bits,2 CPUs LE) LE CPU Freq: 1180 1182 1184 1184 1184 1184 1176 1187 1187 RAM size: 967 MB, # CPU hardware threads: 2 RAM usage: 441 MB, # Benchmark threads: 2 Compressing | Decompressing Dict Speed Usage R/U Rating | Speed Usage R/U Rating KiB/s % MIPS MIPS | KiB/s % MIPS MIPS 22: 621 144 419 605 | 21441 194 942 1831 23: 661 158 426 674 7-Zip (a) [32] 16.02 : Copyright (c) 1999-2016 Igor Pavlov : 2016-05-21 p7zip Version 16.02 (locale=en_US.UTF-8,Utf16=on,HugeFiles=on,32 bits,2 CPUs LE) LE CPU Freq: 1184 1187 1188 1187 1182 1187 1187 1187 1187 RAM size: 967 MB, # CPU hardware threads: 2 RAM usage: 441 MB, # Benchmark threads: 2 Compressing | Decompressing Dict Speed Usage R/U Rating | Speed Usage R/U Rating KiB/s % MIPS MIPS | KiB/s % MIPS MIPS 22: 684 158 422 666 | 20419 185 945 1743 23: 661 158 426 674 | 21341 198 932 1847 24: 643 160 432 692 | 19039 181 921 1671 25: 637 165 441 728 Compression: 698 Decompression: 1695 Total: 1196 ########################################################################## Testing clockspeeds again. System health now: Time CPU load %cpu %sys %usr %nice %io %irq Temp 06:49:33: 1200MHz 2.73 93% 9% 84% 0% 0% 0% 61.9°C Checking cpufreq OPP: Cpufreq OPP: 1200 Measured: 1198.119/1198.383/1198.341 Cpufreq OPP: 1152 Measured: 1150.163/1150.329/1150.291 Cpufreq OPP: 1104 Measured: 1097.800/651.935/723.877 Cpufreq OPP: 1056 Measured: 235.539/1295.718/1041.458 Cpufreq OPP: 1008 Measured: 852.545/852.523/850.844 Cpufreq OPP: 960 Measured: 814.583/816.192/709.484 Cpufreq OPP: 912 Measured: 808.853/953.517/905.895 Cpufreq OPP: 864 Measured: 853.779/859.301/856.789 Cpufreq OPP: 720 Measured: 2672.244/825.888/712.282 Cpufreq OPP: 528 Measured: 301.890/323.499/321.582 Cpufreq OPP: 312 Measured: 287.110/140.695/156.816 Cpufreq OPP: 144 Measured: 141.922/141.875/141.631 ########################################################################## System health while running tinymembench: Time CPU load %cpu %sys %usr %nice %io %irq Temp 06:24:49: 1200MHz 1.47 22% 7% 12% 0% 2% 0% 56.0°C 06:26:50: 1200MHz 1.61 65% 10% 52% 0% 3% 0% 57.9°C 06:28:51: 1200MHz 1.77 57% 7% 49% 0% 0% 0% 58.3°C 06:30:52: 1200MHz 1.48 58% 7% 50% 0% 0% 0% 59.1°C 06:32:53: 1200MHz 1.56 57% 7% 50% 0% 0% 0% 58.0°C 06:34:54: 1200MHz 1.31 57% 6% 50% 0% 0% 0% 58.0°C 06:36:56: 1200MHz 1.42 56% 7% 48% 0% 0% 0% 57.5°C 06:38:56: 1200MHz 1.63 57% 7% 50% 0% 0% 0% 57.9°C System health while running OpenSSL benchmark: Time CPU load %cpu %sys %usr %nice %io %irq Temp 06:40:16: 1200MHz 1.55 31% 7% 21% 0% 1% 0% 57.7°C 06:40:26: 1200MHz 1.72 54% 6% 47% 0% 0% 0% 58.3°C 06:40:37: 1200MHz 1.84 54% 7% 47% 0% 0% 0% 58.4°C 06:40:48: 1200MHz 1.71 54% 6% 47% 0% 0% 0% 58.4°C 06:40:58: 1200MHz 1.76 54% 6% 47% 0% 0% 0% 58.8°C 06:41:09: 1200MHz 1.64 54% 7% 47% 0% 0% 0% 58.7°C 06:41:20: 1200MHz 1.70 54% 6% 47% 0% 0% 0% 58.8°C 06:41:30: 1200MHz 1.90 55% 6% 47% 0% 0% 0% 58.7°C 06:41:41: 1200MHz 1.83 57% 6% 50% 0% 0% 0% 58.8°C 06:41:52: 1200MHz 1.86 57% 7% 50% 0% 0% 0% 58.8°C 06:42:02: 1200MHz 1.73 57% 6% 50% 0% 0% 0% 58.9°C System health while running 7-zip single core benchmark: Time CPU load %cpu %sys %usr %nice %io %irq Temp 06:42:05: 1200MHz 1.73 32% 7% 22% 0% 1% 0% 58.7°C 06:43:06: 1200MHz 2.17 57% 8% 49% 0% 0% 0% 58.7°C 06:44:08: 1200MHz 2.15 57% 8% 48% 0% 0% 0% 58.7°C System health while running 7-zip multi core benchmark: Time CPU load %cpu %sys %usr %nice %io %irq Temp 06:45:10: 1200MHz 2.19 33% 7% 23% 0% 1% 0% 58.5°C 06:45:32: 1200MHz 2.37 88% 8% 79% 0% 0% 0% 60.1°C 06:45:52: 1200MHz 2.29 87% 9% 77% 0% 0% 0% 60.5°C 06:46:14: 1200MHz 2.50 91% 9% 81% 0% 0% 0% 61.1°C 06:46:36: 1200MHz 2.71 87% 9% 78% 0% 0% 0% 61.0°C 06:46:57: 1200MHz 2.73 92% 9% 82% 0% 0% 0% 61.3°C 06:47:22: 1200MHz 2.60 87% 8% 79% 0% 0% 0% 62.0°C 06:47:43: 1200MHz 2.42 82% 9% 72% 0% 0% 0% 61.5°C 06:48:08: 1200MHz 2.23 91% 8% 82% 0% 0% 0% 62.4°C 06:48:30: 1200MHz 2.54 85% 9% 75% 0% 0% 0% 62.1°C 06:48:50: 1200MHz 2.42 88% 9% 78% 0% 0% 0% 62.0°C 06:49:11: 1200MHz 2.70 89% 9% 80% 0% 0% 0% 62.2°C 06:49:33: 1200MHz 2.73 93% 9% 84% 0% 0% 0% 61.9°C ########################################################################## Linux 5.8.16-sunxi (bananapi) 12/06/2020 _armv7l_ (2 CPU) avg-cpu: %user %nice %system %iowait %steal %idle 27.27 0.02 7.94 1.49 0.00 63.29 Device tps kB_read/s kB_wrtn/s kB_read kB_wrtn mmcblk0 1.63 53.32 37.44 232088 162968 zram0 0.10 0.43 0.52 1880 2248 zram1 0.07 0.27 0.00 1184 4 total used free shared buff/cache available Mem: 967Mi 128Mi 552Mi 2.0Mi 285Mi 805Mi Swap: 483Mi 0B 483Mi Filename Type Size Used Priority /dev/zram1 partition 495108 0 5 Architecture: armv7l Byte Order: Little Endian CPU(s): 2 On-line CPU(s) list: 0,1 Thread(s) per core: 1 Core(s) per socket: 2 Socket(s): 1 Vendor ID: ARM Model: 4 Model name: Cortex-A7 Stepping: r0p4 CPU max MHz: 1200.0000 CPU min MHz: 144.0000 BogoMIPS: 63.15 Flags: half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm