Oracle Cloud 無料枠VMのUnixBenchを測った

Oracleが提供しているクラウドサービス Oracle Cloud Infrastructure (OCI) には無料でLinux VM(コンピュート)が使える無料枠があります。

今回、メモリ1GBでCPUがAMDなVM(VM.Standard.E2.1.Micro)を作成できたので、UnixBenchでスコアを測ってみました。

OCI (VM.Standard.E2.1.Micro)

1コアのスコアは525.7、マルチコアのスコアは525.7でした。

========================================================================        
   BYTE UNIX Benchmarks (Version 5.1.3)                                         
                               
   System: ***: GNU/Linux                                         
   OS: GNU/Linux -- 5.11.0-1022-oracle -- #23~20.04.1-Ubuntu SMP Fri Nov 12 15:45:30 UTC 2021
   Machine: x86_64 (x86_64)                                           
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")            
   CPU 0: AMD EPYC 7551 32-Core Processor (3992.5 bogomips)           
          Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
   CPU 1: AMD EPYC 7551 32-Core Processor (3992.5 bogomips)           
          Hyper-Threading, x86-64, MMX, AMD MMX, Physical Address Ext, SYSENTER/SYSEXIT, AMD virtualization, SYSCALL/SYSRET
   09:42:36 up 43 min,  2 users,  load average: 0.21, 0.44, 0.24; runlevel 2022-01-29



------------------------------------------------------------------------        
Benchmark Run: Sat Jan 29 2022 09:42:36 - 10:10:30     
2 CPUs in system; running 1 parallel copy of tests     

Dhrystone 2 using register variables       14824047.7 lps   (10.0 s, 7 samples) 
Double-Precision Whetstone                     2492.9 MWIPS (9.7 s, 7 samples)  
Execl Throughput                               1581.4 lps   (30.0 s, 2 samples) 
File Copy 1024 bufsize 2000 maxblocks        371014.0 KBps  (30.0 s, 2 samples) 
File Copy 256 bufsize 500 maxblocks          102532.9 KBps  (30.1 s, 2 samples) 
File Copy 4096 bufsize 8000 maxblocks        964912.8 KBps  (30.0 s, 2 samples) 
Pipe Throughput                              656424.9 lps   (10.0 s, 7 samples) 
Pipe-based Context Switching                  30052.0 lps   (10.0 s, 7 samples) 
Process Creation                               2187.2 lps   (30.0 s, 2 samples) 
Shell Scripts (1 concurrent)                   3050.5 lpm   (60.0 s, 2 samples) 
Shell Scripts (8 concurrent)                    411.0 lpm   (60.1 s, 2 samples) 
System Call Overhead                         964589.4 lps   (10.0 s, 7 samples) 

System Benchmarks Index Values               BASELINE       RESULT    INDEX     
Dhrystone 2 using register variables         116700.0   14824047.7   1270.3     
Double-Precision Whetstone                       55.0       2492.9    453.3     
Execl Throughput                                 43.0       1581.4    367.8     
File Copy 1024 bufsize 2000 maxblocks          3960.0     371014.0    936.9     
File Copy 256 bufsize 500 maxblocks            1655.0     102532.9    619.5     
File Copy 4096 bufsize 8000 maxblocks          5800.0     964912.8   1663.6     
Pipe Throughput                               12440.0     656424.9    527.7     
Pipe-based Context Switching                   4000.0      30052.0     75.1     
Process Creation                                126.0       2187.2    173.6     
Shell Scripts (1 concurrent)                     42.4       3050.5    719.5     
Shell Scripts (8 concurrent)                      6.0        411.0    685.0     
System Call Overhead                          15000.0     964589.4    643.1     
                   ========     
System Benchmarks Index Score                                         525.7     

------------------------------------------------------------------------        
Benchmark Run: Sat Jan 29 2022 10:10:30 - 10:38:33     
2 CPUs in system; running 2 parallel copies of tests   

Dhrystone 2 using register variables       15153003.2 lps   (10.0 s, 7 samples) 
Double-Precision Whetstone                     2532.4 MWIPS (9.8 s, 7 samples)  
Execl Throughput                               1528.6 lps   (29.7 s, 2 samples) 
File Copy 1024 bufsize 2000 maxblocks        267085.8 KBps  (30.0 s, 2 samples) 
File Copy 256 bufsize 500 maxblocks           57288.8 KBps  (30.0 s, 2 samples) 
File Copy 4096 bufsize 8000 maxblocks        682638.1 KBps  (30.0 s, 2 samples) 
Pipe Throughput                              665793.1 lps   (10.0 s, 7 samples) 
Pipe-based Context Switching                  86892.0 lps   (10.0 s, 7 samples) 
Process Creation                               3993.7 lps   (30.0 s, 2 samples) 
Shell Scripts (1 concurrent)                   3186.1 lpm   (60.0 s, 2 samples) 
Shell Scripts (8 concurrent)                    412.2 lpm   (60.2 s, 2 samples) 
System Call Overhead                         795509.2 lps   (10.0 s, 7 samples) 

System Benchmarks Index Values               BASELINE       RESULT    INDEX     
Dhrystone 2 using register variables         116700.0   15153003.2   1298.5     
Double-Precision Whetstone                       55.0       2532.4    460.4     
Execl Throughput                                 43.0       1528.6    355.5     
File Copy 1024 bufsize 2000 maxblocks          3960.0     267085.8    674.5     
File Copy 256 bufsize 500 maxblocks            1655.0      57288.8    346.2     
File Copy 4096 bufsize 8000 maxblocks          5800.0     682638.1   1177.0     
Pipe Throughput                               12440.0     665793.1    535.2     
Pipe-based Context Switching                   4000.0      86892.0    217.2     
Process Creation                                126.0       3993.7    317.0     
Shell Scripts (1 concurrent)                     42.4       3186.1    751.4     
Shell Scripts (8 concurrent)                      6.0        412.2    687.0     
System Call Overhead                          15000.0     795509.2    530.3     
                   ========     
System Benchmarks Index Score                                         538.1

GCP (e2-micro)

OCIだけでは、比較にならないのでGCP無料枠で使えるVM(e2-micro)のスコアも測ってみました。
1コアのスコアは136.4、マルチコアのスコアは187.8でした。OCIの方が3.8倍高いですね。

========================================================================        
   BYTE UNIX Benchmarks (Version 5.1.3)                   
                                                          
   System: ***: GNU/Linux                            
   OS: GNU/Linux -- 5.11.0-1017-gcp -- #19~20.04.1-Ubuntu SMP Thu Aug 12 05:25:25 UTC 2021
   Machine: x86_64 (x86_64)                               
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: Intel(R) Xeon(R) CPU @ 2.20GHz (4400.0 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET                                                                           
   CPU 1: Intel(R) Xeon(R) CPU @ 2.20GHz (4400.0 bogomips)
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET                                                                           
   10:21:30 up 7 min,  2 users,  load average: 0.25, 0.67, 0.47; runlevel 2022-01-29      
                                                          
------------------------------------------------------------------------        
Benchmark Run: Sat Jan 29 2022 10:21:30 - 10:51:19        
2 CPUs in system; running 1 parallel copy of tests        

Dhrystone 2 using register variables       13641120.9 lps   (10.0 s, 7 samples) 
Double-Precision Whetstone                      713.2 MWIPS (13.8 s, 7 samples) 
Execl Throughput                                418.4 lps   (30.0 s, 2 samples) 
File Copy 1024 bufsize 2000 maxblocks         64182.0 KBps  (30.1 s, 2 samples) 
File Copy 256 bufsize 500 maxblocks           18290.5 KBps  (30.1 s, 2 samples) 
File Copy 4096 bufsize 8000 maxblocks        157252.3 KBps  (30.1 s, 2 samples) 
Pipe Throughput                              131333.0 lps   (10.1 s, 7 samples) 
Pipe-based Context Switching                  13275.6 lps   (10.1 s, 7 samples) 
Process Creation                                735.0 lps   (30.0 s, 2 samples) 
Shell Scripts (1 concurrent)                    952.1 lpm   (60.2 s, 2 samples) 
Shell Scripts (8 concurrent)                    147.5 lpm   (60.2 s, 2 samples) 
System Call Overhead                          77329.0 lps   (10.1 s, 7 samples) 

System Benchmarks Index Values               BASELINE       RESULT    INDEX         
Dhrystone 2 using register variables         116700.0   13641120.9   1168.9         
Double-Precision Whetstone                       55.0        713.2    129.7         
Execl Throughput                                 43.0        418.4     97.3         
File Copy 1024 bufsize 2000 maxblocks          3960.0      64182.0    162.1         
File Copy 256 bufsize 500 maxblocks            1655.0      18290.5    110.5         
File Copy 4096 bufsize 8000 maxblocks          5800.0     157252.3    271.1         
Pipe Throughput                               12440.0     131333.0    105.6         
Pipe-based Context Switching                   4000.0      13275.6     33.2         
Process Creation                                126.0        735.0     58.3         
Shell Scripts (1 concurrent)                     42.4        952.1    224.5         
Shell Scripts (8 concurrent)                      6.0        147.5    245.8         
System Call Overhead                          15000.0      77329.0     51.6         
                                                                   ========         
System Benchmarks Index Score                                         136.4  
------------------------------------------------------------------------
Benchmark Run: Sat Jan 29 2022 10:51:19 - 11:20:46
2 CPUs in system; running 2 parallel copies of tests

Dhrystone 2 using register variables        8992568.5 lps   (10.1 s, 7 samples)
Double-Precision Whetstone                     1257.5 MWIPS (12.8 s, 7 samples)
Execl Throughput                                580.2 lps   (29.8 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks         83211.3 KBps  (30.1 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           22899.4 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        222590.0 KBps  (30.1 s, 2 samples)
Pipe Throughput                              171582.3 lps   (10.1 s, 7 samples)
Pipe-based Context Switching                  38997.6 lps   (10.1 s, 7 samples)
Process Creation                               1556.3 lps   (30.1 s, 2 samples)
Shell Scripts (1 concurrent)                   1135.5 lpm   (60.1 s, 2 samples)
Shell Scripts (8 concurrent)                    148.4 lpm   (60.4 s, 2 samples)
System Call Overhead                          99051.6 lps   (10.1 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0    8992568.5    770.6
Double-Precision Whetstone                       55.0       1257.5    228.6
Execl Throughput                                 43.0        580.2    134.9
File Copy 1024 bufsize 2000 maxblocks          3960.0      83211.3    210.1
File Copy 256 bufsize 500 maxblocks            1655.0      22899.4    138.4
File Copy 4096 bufsize 8000 maxblocks          5800.0     222590.0    383.8
Pipe Throughput                               12440.0     171582.3    137.9
Pipe-based Context Switching                   4000.0      38997.6     97.5
Process Creation                                126.0       1556.3    123.5
Shell Scripts (1 concurrent)                     42.4       1135.5    267.8
Shell Scripts (8 concurrent)                      6.0        148.4    247.3
System Call Overhead                          15000.0      99051.6     66.0
                                                                   ========
System Benchmarks Index Score                                         187.8


real    59m21.228s
user    37m1.455s
sys     36m35.735s

Indigo (2GBメモリ)

今契約しているVPS WebArena Indigo(CPU2コアメモリ2GB)のUnixBenchも測ってみました。
1コアのスコアは583.3、マルチコアのスコアは1047.6でした。1コアはOCIとどっこいですね。

========================================================================
   BYTE UNIX Benchmarks (Version 5.1.3)                                 
                                                                        
   System: pl-210804: GNU/Linux                                         
   OS: GNU/Linux -- 5.4.0-91-generic -- #102-Ubuntu SMP Fri Nov 5 16:31:28 UTC 2021           
   Machine: x86_64 (x86_64)                                             
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")              
   CPU 0: Intel Xeon E312xx (Sandy Bridge) (4400.0 bogomips)            
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   CPU 1: Intel Xeon E312xx (Sandy Bridge) (4400.0 bogomips)            
          Hyper-Threading, x86-64, MMX, Physical Address Ext, SYSENTER/SYSEXIT, SYSCALL/SYSRET
   09:49:29 up 48 days,  5:22,  2 users,  load average: 0.63, 0.35, 0.13; runlevel 2021-12-12 



------------------------------------------------------------------------
Benchmark Run: Sat Jan 29 2022 09:49:29 - 10:17:31
2 CPUs in system; running 1 parallel copy of tests

Dhrystone 2 using register variables       29511002.9 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     4329.2 MWIPS (9.9 s, 7 samples)
Execl Throughput                               2495.0 lps   (30.0 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        246087.4 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks           71250.8 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks        619403.4 KBps  (30.0 s, 2 samples)
Pipe Throughput                              403720.8 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                  48445.0 lps   (10.0 s, 7 samples)
Process Creation                               4570.7 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   5031.1 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                    952.9 lpm   (60.0 s, 2 samples)
System Call Overhead                         262527.2 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   29511002.9   2528.8
Double-Precision Whetstone                       55.0       4329.2    787.1
Execl Throughput                                 43.0       2495.0    580.2
File Copy 1024 bufsize 2000 maxblocks          3960.0     246087.4    621.4
File Copy 256 bufsize 500 maxblocks            1655.0      71250.8    430.5
File Copy 4096 bufsize 8000 maxblocks          5800.0     619403.4   1067.9
Pipe Throughput                               12440.0     403720.8    324.5
Pipe-based Context Switching                   4000.0      48445.0    121.1
Process Creation                                126.0       4570.7    362.8
Shell Scripts (1 concurrent)                     42.4       5031.1   1186.6
Shell Scripts (8 concurrent)                      6.0        952.9   1588.1
System Call Overhead                          15000.0     262527.2    175.0
                                                                   ========
System Benchmarks Index Score                                         583.3

------------------------------------------------------------------------
Benchmark Run: Sat Jan 29 2022 10:17:31 - 10:45:42
2 CPUs in system; running 2 parallel copies of tests

Dhrystone 2 using register variables       56890710.9 lps   (10.0 s, 7 samples)
Double-Precision Whetstone                     8459.4 MWIPS (9.9 s, 7 samples)
Execl Throughput                               4053.0 lps   (29.6 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks        432408.4 KBps  (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks          122486.4 KBps  (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks       1087348.2 KBps  (30.0 s, 2 samples)
Pipe Throughput                              791083.5 lps   (10.0 s, 7 samples)
Pipe-based Context Switching                 157821.7 lps   (10.0 s, 7 samples)
Process Creation                               9129.5 lps   (30.0 s, 2 samples)
Shell Scripts (1 concurrent)                   6818.8 lpm   (60.0 s, 2 samples)
Shell Scripts (8 concurrent)                   1015.6 lpm   (60.1 s, 2 samples)
System Call Overhead                         494552.3 lps   (10.0 s, 7 samples)

System Benchmarks Index Values               BASELINE       RESULT    INDEX
Dhrystone 2 using register variables         116700.0   56890710.9   4875.0
Double-Precision Whetstone                       55.0       8459.4   1538.1
Execl Throughput                                 43.0       4053.0    942.6
File Copy 1024 bufsize 2000 maxblocks          3960.0     432408.4   1091.9
File Copy 256 bufsize 500 maxblocks            1655.0     122486.4    740.1
File Copy 4096 bufsize 8000 maxblocks          5800.0    1087348.2   1874.7
Pipe Throughput                               12440.0     791083.5    635.9
Pipe-based Context Switching                   4000.0     157821.7    394.6
Process Creation                                126.0       9129.5    724.6
Shell Scripts (1 concurrent)                     42.4       6818.8   1608.2
Shell Scripts (8 concurrent)                      6.0       1015.6   1692.6
System Call Overhead                          15000.0     494552.3    329.7
                                                                   ========
System Benchmarks Index Score                                        1047.6

感想

OCIの無料枠VM、意外と性能良いですね。今契約しているVPSを解約して、移ってもいいかもしれません。
学習環境としても悪くないですね。
(Ubuntuはポートを解放するためにiptablesを使う必要があったりと、クセはあるようですが。)
ただ、日本リージョン(大阪)だと、人気なのか在庫不足で無料枠でVMが作成できないため、海外リージョンを狙った方が良さそうです。

関連記事