忍者ブログ

one's way

個人の趣味で色々な物を触ったり、検証した内容等纏めたブログです。

【Unix】unixbenchの備忘録

最終更新日:2019/05/19

先日購入したラズパイ3B+でベンチマークを動かして見たときのまとめです。
尚、使用したベンチマークですが、unixbenchを使用してみました。
今度PC(Ubuntu)でもやってみようと思うので、その為の備忘録です。



0.目次



1.検証対象
2.unixbenchの導入・実行手順
3.実行結果
3-1)ラズパイ3B+


1.検証対象



  ①Raspberry Pi 3B+(以下ラズパイ3B+)
   CPU:
Broadcom BCM2837B0, Cortex-A53 (ARMv8)1.4GHz
   MEM:1GB
   OS :raspberrypi 4.14.98-v7+
   備考 :MicroSDは"SAMSUNG EVO 32GB Class10 UHS-I"

  ※今後追加予定


2.unixbenchの導入・実行手順



以下のコマンドを順次実行する
sudo apt install git
git clone https://github.com/kdlucas/byte-unixbench
cd byte-unixbench/UnixBench
./Run



3.実行結果


3-1)ラズパイ3B+

○実行結果
===========================================================
   BYTE UNIX Benchmarks (Version 5.1.3)

   System: raspberrypi: GNU/Linux
   OS: GNU/Linux -- 4.14.98-v7+ -- #1200 SMP Tue Feb 12 20:27:48 GMT 2019
   Machine: armv7l (unknown)
   Language: en_US.utf8 (charmap="ANSI_X3.4-1968", collate="ANSI_X3.4-1968")
   CPU 0: ARMv7 Processor rev 4 (v7l) (0.0 bogomips)

   CPU 1: ARMv7 Processor rev 4 (v7l) (0.0 bogomips)

   CPU 2: ARMv7 Processor rev 4 (v7l) (0.0 bogomips)

   CPU 3: ARMv7 Processor rev 4 (v7l) (0.0 bogomips)

   11:27:46 up 8 min,  3 users,  load average: 0.19, 0.07, 0.03; runlevel May

------------------------------------------------------------------------
Benchmark Run: 木  5月 16 2019 19:27:46 - 19:55:55
4 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables        5020081.5 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     1153.8 MWIPS (10.5 s, 7 samples)
Execl Throughput                                912.1 lps   (29.8 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        132906.9 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           38249.5 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        324672.4 KBps  (30.0 s, 2 samples)
Pipe Throughput                              252918.9 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  53314.3 lps   (10.0 s, 7 samples)
Process Creation                               2283.6 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   1948.7 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    585.4 lpm   (60.1 s, 2 samples)
System Call Overhead                         600047.6 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0    5020081.5    430.2
Double-Precision Whetstone                       55.0       1153.8    209.8
Execl Throughput                                 43.0        912.1    212.1
File Copy 1024 bufsize 2000 maxblocks          3960.0     132906.9    335.6
File Copy 256 bufsize 500 maxblocks            1655.0      38249.5    231.1
File Copy 4096 bufsize 8000 maxblocks          5800.0     324672.4    559.8
Pipe Throughput                               12440.0     252918.9    203.3
Pipe-based Context Switching                   4000.0      53314.3    133.3
Process Creation                                126.0       2283.6    181.2
Shell Scripts (1 concurrent)                     42.4       1948.7    459.6
Shell Scripts (8 concurrent)                      6.0        585.4    975.7
System Call Overhead                          15000.0     600047.6    400.0
                                                                   ========
System Benchmarks Index Score                                         308.1

------------------------------------------------------------------------
Benchmark Run: 木  5月 16 2019 19:55:55 - 20:24:17
4 CPUs in system; running 4 parallel copies of tests

Dhrystone 2 using register variables       17260653.9 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     4215.3 MWIPS (9.9 s, 7 samples)
Execl Throughput                               2124.1 lps   (29.7 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        212212.6 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           59000.9 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        555417.9 KBps  (30.0 s, 2 samples)
Pipe Throughput                             1006072.5 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 222278.1 lps   (10.0 s, 7 samples)
Process Creation                               4968.5 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   4389.9 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    609.6 lpm   (60.2 s, 2 samples)
System Call Overhead                        2307906.7 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   17260653.9   1479.1
Double-Precision Whetstone                       55.0       4215.3    766.4
Execl Throughput                                 43.0       2124.1    494.0
File Copy 1024 bufsize 2000 maxblocks          3960.0     212212.6    535.9
File Copy 256 bufsize 500 maxblocks            1655.0      59000.9    356.5
File Copy 4096 bufsize 8000 maxblocks          5800.0     555417.9    957.6
Pipe Throughput                               12440.0    1006072.5    808.7
Pipe-based Context Switching                   4000.0     222278.1    555.7
Process Creation                                126.0       4968.5    394.3
Shell Scripts (1 concurrent)                     42.4       4389.9   1035.4
Shell Scripts (8 concurrent)                      6.0        609.6   1016.0
System Call Overhead                          15000.0    2307906.7   1538.6
                                                                   ========
System Benchmarks Index Score                                         745.3

○ベンチ結果まとめ
 ・スコア :1 parallel(308.1)/4 parallel(745.3)
 ・実行時間:1 parallel(28分09秒)/4 parallel(28分22秒)/合計(56分31秒)
 ・最高温度:4 parallelの"4 x Shell Scripts (8 concurrent) "中にtemp=80.1'Cを記録
  ※室温24度
 ・特記事項:実行中に幾つかスプリプトのエラーが出ていた


とりあえず次はジャンクPCのVersaproでも走らせてみる予定です。

○更新履歴

以上

タグもどき:unix ラズパイ Bench Mark
PR