Rock Pi S 届いた。

Rock Pi S スペック概要

CPU	Cortex-A35 x4
MEM	256MB / 512MB
LAN	100Mbps
サイズ	43.2㎜ x 43.2㎜

販売サイト
https://shop.allnetchina.cn/products/rock-pi-s
※ 38㎜ x 38mmと記載ありますが、間違いです!!

起動までの手順

OSイメージ

Debian落としてSDに焼く。

UART (Serial Console)

baudrate: 1500000
data bit: 8
stop bit: 1
parity : none

配線の場所は、RaspPiと同じ#8, #10 のところ。

Debian 起動

User Name : rock
Password : rock

Linux側の情報

uname -a

Linux rockpis 4.4.143-23-rockchip-g394d86b2b38b #1 SMP PREEMPT Mon Aug 5 12:15:39 UTC 2019 aarch64 GNU/Linux

df

rock@rockpis:~$ df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/root 14740444 893308 13220600 7% /
devtmpfs 217960 0 217960 0% /dev
tmpfs 218152 0 218152 0% /dev/shm
tmpfs 218152 5740 212412 3% /run
tmpfs 5120 4 5116 1% /run/lock
tmpfs 218152 0 218152 0% /sys/fs/cgroup
tmpfs 43628 0 43628 0% /run/user/1000

cat /proc/cpuinfo

rock@rockpis:~$ cat /proc/cpuinfo
processor : 0
BogoMIPS : 48.00
Features : fp asimd aes pmull sha1 sha2 crc32
CPU implementer : 0x41
CPU architecture: 8
CPU variant : 0x0
CPU part : 0xd04
CPU revision : 2

processor : 1
BogoMIPS : 48.00
Features : fp asimd aes pmull sha1 sha2 crc32
CPU implementer : 0x41
CPU architecture: 8
CPU variant : 0x0
CPU part : 0xd04
CPU revision : 2

processor : 2
BogoMIPS : 48.00
Features : fp asimd aes pmull sha1 sha2 crc32
CPU implementer : 0x41
CPU architecture: 8
CPU variant : 0x0
CPU part : 0xd04
CPU revision : 2

processor : 3
BogoMIPS : 48.00
Features : fp asimd aes pmull sha1 sha2 crc32
CPU implementer : 0x41
CPU architecture: 8
CPU variant : 0x0
CPU part : 0xd04
CPU revision : 2

u-boot あたりのログ

U-Boot 2017.09-02368-ga5dffc8 (Aug 05 2019 - 11:44:53 +0800)

Model: Radxa ROCK Pi S
PreSerial: 0
DRAM: 510 MiB
Relocation Offset is: 1f92c000
vdd_core 1044000 uV
vdd-core init 1044000 uV
MMC: dwmmc@ff480000: 0, dwmmc@ff490000: 1
Using default environment

消費電流

計測モデルは、
MEM: 512MB
Wifi: なし
NAND: なし

	消費電流
LAN接続なし、アイドリング時	38 mA
LAN接続あり、アイドリング時	68 mA

NanoPi NEO2やRaspPiZero のアイドリング時は100mAを超えるのでかなり少ない。

起動時

アイドリング時

演算速度

いつもので実施。

4x4行列同士の掛算を高速化してみる ～Part2～

実施結果

Function name	Single Thread	Multi Thread	Performance(MT)	MT/ST
MultMatrix_ForC	4.4 M mul/s	17.8 M mul/s	1.991 Gflops	400 %
MultMatrix_ForR	2.2 M mul/s	8.8 M mul/s	0.990 Gflops	400 %
MultMatrix_ExpC	2.5 M mul/s	9.9 M mul/s	1.106 Gflops	400 %
MultMatrix_ExpR	2.5 M mul/s	9.8 M mul/s	1.102 Gflops	399 %
MultMatrixF_NEON	8.8 M mul/s	35.1 M mul/s	3.927 Gflops	400 %
MultMatrixF_NEON_type2	7.9 M mul/s	31.5 M mul/s	3.526 Gflops	400 %
MultMatrix_ForC	2.4 M mul/s	9.8 M mul/s	1.093 Gflops	400 %
MultMatrix_ForR	2.2 M mul/s	8.8 M mul/s	0.989 Gflops	400 %
MultMatrix_ExpC	2.5 M mul/s	9.8 M mul/s	1.103 Gflops	400 %
MultMatrix_ExpR	2.5 M mul/s	9.8 M mul/s	1.103 Gflops	400 %
CalcInverse	1.7 M inv/s	6.9 M inv/s	1.694 Gflops	400 %
CalcInverse	1.4 M inv/s	5.5 M inv/s	1.355 Gflops	400 %
CopyMatrix_memcpy	17.8 M cpy/s	71.3 M cpy/s	9.121 GB/sec	400 %
CopyMatrix_Expand	19.2 M cpy/s	76.8 M cpy/s	9.828 GB/sec	400 %
CopyMatrixD_NEON	27.7 M cpy/s	110.8 M cpy/s	14.187 GB/sec	400 %

消費電流

Cortex-A53 x4 のNanoPi NEO2 にくらべて 60%~90% 程度。
消費電流は、ピークで180mA程度。

NEO2 の場合 1200mAくらいまで上がるので、かなり電力効率は良いね！

UnixBench

実施コマンド

$ wget https://storage.googleapis.com/google-code-archive-downloads/v2/code.google.com/byte-unixbench/UnixBench5.1.3.tgz
$ tar -zxvf UnixBench5.1.3.tgz
$ cd ./UnixBench
$ ./Run -c 1 -c 4

結果

result

   #    #  #    #  #  #    #          #####   ######  #    #   ####   #    #
   #    #  ##   #  #   #  #           #    #  #       ##   #  #    #  #    #
   #    #  # #  #  #    ##            #####   #####   # #  #  #       ######
   #    #  #  # #  #    ##            #    #  #       #  # #  #       #    #
   #    #  #   ##  #   #  #           #    #  #       #   ##  #    #  #    #
    ####   #    #  #  #    #          #####   ######  #    #   ####   #    #

   Version 5.1.3                      Based on the Byte Magazine Unix Benchmark

   Multi-CPU version                  Version 5 revisions by Ian Smith,
                                      Sunnyvale, CA, USA
   January 13, 2011                   johantheghost at yahoo period com


1 x Dhrystone 2 using register variables  1 2 3 4 5 6 7 8 9 10

1 x Double-Precision Whetstone  1 2 3 4 5 6 7 8 9 10

1 x Execl Throughput  1 2 3

1 x File Copy 1024 bufsize 2000 maxblocks  1 2 3

1 x File Copy 256 bufsize 500 maxblocks  1 2 3

1 x File Copy 4096 bufsize 8000 maxblocks  1 2 3

1 x Pipe Throughput  1 2 3 4 5 6 7 8 9 10

1 x Pipe-based Context Switching  1 2 3 4 5 6 7 8 9 10

1 x Process Creation  1 2 3

1 x System Call Overhead  1 2 3 4 5 6 7 8 9 10

1 x Shell Scripts (1 concurrent)  1 2 3

1 x Shell Scripts (8 concurrent)  1 2 3

4 x Dhrystone 2 using register variables  1 2 3 4 5 6 7 8 9 10

4 x Double-Precision Whetstone  1 2 3 4 5 6 7 8 9 10

4 x Execl Throughput  1 2 3

4 x File Copy 1024 bufsize 2000 maxblocks  1 2 3

4 x File Copy 256 bufsize 500 maxblocks  1 2 3

4 x File Copy 4096 bufsize 8000 maxblocks  1 2 3

4 x Pipe Throughput  1 2 3 4 5 6 7 8 9 10

4 x Pipe-based Context Switching  1 2 3 4 5 6 7 8 9 10

4 x Process Creation  1 2 3

4 x System Call Overhead  1 2 3 4 5 6 7 8 9 10

4 x Shell Scripts (1 concurrent)  1 2 3

4 x Shell Scripts (8 concurrent)  1 2 3

========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: rockpis: GNU/Linux
   OS: GNU/Linux -- 4.4.143-26-rockchip-g5dbd30f3e92b -- #1 SMP PREEMPT Sat Sep 7 02:56:45 UTC 2019
   Machine: aarch64 (unknown)
   Language: en_US.utf8 (charmap="ANSI_X3.4-1968", collate="ANSI_X3.4-1968")
   07:11:05 up 2 min,  1 user,  load average: 0.50, 0.20, 0.08; runlevel 5

------------------------------------------------------------------------
Benchmark Run: Mon Sep 09 2019 07:11:05 - 07:39:29
0 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables        3987296.7 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                      838.2 MWIPS (10.0 s, 7 samples)
Execl Throughput                                635.4 lps   (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        104064.5 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           30723.5 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        252105.3 KBps  (30.0 s, 2 samples)
Pipe Throughput                              316565.5 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  16788.3 lps   (10.0 s, 7 samples)
Process Creation                                742.1 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   1529.5 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    454.0 lpm   (60.1 s, 2 samples)
System Call Overhead                         532097.5 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0    3987296.7    341.7
Double-Precision Whetstone                       55.0        838.2    152.4
Execl Throughput                                 43.0        635.4    147.8
File Copy 1024 bufsize 2000 maxblocks          3960.0     104064.5    262.8
File Copy 256 bufsize 500 maxblocks            1655.0      30723.5    185.6
File Copy 4096 bufsize 8000 maxblocks          5800.0     252105.3    434.7
Pipe Throughput                               12440.0     316565.5    254.5
Pipe-based Context Switching                   4000.0      16788.3     42.0
Process Creation                                126.0        742.1     58.9
Shell Scripts (1 concurrent)                     42.4       1529.5    360.7
Shell Scripts (8 concurrent)                      6.0        454.0    756.7
System Call Overhead                          15000.0     532097.5    354.7
                                                                   ========
System Benchmarks Index Score                                         215.3

------------------------------------------------------------------------
Benchmark Run: Mon Sep 09 2019 07:39:29 - 08:07:51
0 CPUs in system; running 4 parallel copies of tests

Dhrystone 2 using register variables       15967104.2 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     3356.6 MWIPS (10.0 s, 7 samples)
Execl Throughput                               1895.4 lps   (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        156650.9 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           44580.9 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        443744.6 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1253868.3 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 172940.2 lps   (10.0 s, 7 samples)
Process Creation                               3909.1 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   3617.7 lpm   (60.1 s, 2 samples)
Shell Scripts (8 concurrent)                    466.7 lpm   (60.4 s, 2 samples)
System Call Overhead                        2050045.5 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   15967104.2   1368.2
Double-Precision Whetstone                       55.0       3356.6    610.3
Execl Throughput                                 43.0       1895.4    440.8
File Copy 1024 bufsize 2000 maxblocks          3960.0     156650.9    395.6
File Copy 256 bufsize 500 maxblocks            1655.0      44580.9    269.4
File Copy 4096 bufsize 8000 maxblocks          5800.0     443744.6    765.1
Pipe Throughput                               12440.0    1253868.3   1007.9
Pipe-based Context Switching                   4000.0     172940.2    432.4
Process Creation                                126.0       3909.1    310.2
Shell Scripts (1 concurrent)                     42.4       3617.7    853.2
Shell Scripts (8 concurrent)                      6.0        466.7    777.9
System Call Overhead                          15000.0    2050045.5   1366.7
                                                                   ========
System Benchmarks Index Score                                         626.8