葉難: Raspberry Pi 2的效能

只要安裝Raspbian並執行各種應用軟體，就能大概得知Pi 2的硬體效能提昇了多少，CPU從ARMv6到ARMv7、從單核心到四核心，記憶體從512MB到1GB，不消多說也知道變快了，官方號稱執行速度會有六倍之快，不過實際情況還是要看你的程式性質，有無使用多核心、有無運用GPU、等等。

以前若是較大型的應用軟體，例如圖片編輯軟體GIMP、網頁瀏覽器、Scribus、辦公室軟體，光開啟就花去不少時間，使用時反應也非常遲鈍，根本不堪使用，不過現在情況已大為改善，的確可作為個人電腦來用，若再考慮價錢性能比，更是物超所值。

rpi每次都能引爆話題，這次Pi 2也不例外，許多人剛拿到熱騰騰的Pi 2，就迫不及待地進行效能檢測，譬如Ladyada在凌晨三點就以PCSX模擬PlayStation玩遊戲了。

我一拿到Pi 2，就是超頻，預設頻率是900MHz（還是800MHz？），不多說，直接選擇raspi-config超頻選單中的1000MHz。不過重開機、看到登入提示後，居然停住不動了，嘿，檢查之後發現是電源不夠強，更換後便無問題。

然後能以指令cpufreq-info查詢每個核心的運作情況：
pi@raspberrypi ~ $ cpufreq-info
cpufrequtils 008: cpufreq-info (C) Dominik Brodowski 2004-2009
Report errors and bugs to cpufreq@vger.kernel.org, please.
analyzing CPU 0:
driver: BCM2835 CPUFreq
CPUs which run at the same hardware frequency: 0 1 2 3
CPUs which need to have their frequency coordinated by software: 0 1 2 3
maximum transition latency: 355 us.
hardware limits: 600 MHz - 1000 MHz
available frequency steps: 600 MHz, 1000 MHz
available cpufreq governors: conservative, ondemand, userspace, powersave, performance
current policy: frequency should be within 600 MHz and 1000 MHz.
                  The governor "ondemand" may decide which speed to use
                  within this range.
current CPU frequency is 600 MHz.
cpufreq stats: 600 MHz:49.87%, 1000 MHz:50.13% (3)
analyzing CPU 1:
driver: BCM2835 CPUFreq
CPUs which run at the same hardware frequency: 0 1 2 3
CPUs which need to have their frequency coordinated by software: 0 1 2 3
maximum transition latency: 355 us.
hardware limits: 600 MHz - 1000 MHz
available frequency steps: 600 MHz, 1000 MHz
available cpufreq governors: conservative, ondemand, userspace, powersave, performance
current policy: frequency should be within 600 MHz and 1000 MHz.
                  The governor "ondemand" may decide which speed to use
                  within this range.
current CPU frequency is 600 MHz.
cpufreq stats: 600 MHz:49.87%, 1000 MHz:50.13% (3)
analyzing CPU 2:
driver: BCM2835 CPUFreq
CPUs which run at the same hardware frequency: 0 1 2 3
CPUs which need to have their frequency coordinated by software: 0 1 2 3
maximum transition latency: 355 us.
hardware limits: 600 MHz - 1000 MHz
available frequency steps: 600 MHz, 1000 MHz
available cpufreq governors: conservative, ondemand, userspace, powersave, performance
current policy: frequency should be within 600 MHz and 1000 MHz.
                  The governor "ondemand" may decide which speed to use
                  within this range.
current CPU frequency is 600 MHz.
cpufreq stats: 600 MHz:49.87%, 1000 MHz:50.13% (3)
analyzing CPU 3:
driver: BCM2835 CPUFreq
CPUs which run at the same hardware frequency: 0 1 2 3
CPUs which need to have their frequency coordinated by software: 0 1 2 3
maximum transition latency: 355 us.
hardware limits: 600 MHz - 1000 MHz
available frequency steps: 600 MHz, 1000 MHz
available cpufreq governors: conservative, ondemand, userspace, powersave, performance
current policy: frequency should be within 600 MHz and 1000 MHz.
                  The governor "ondemand" may decide which speed to use
                  within this range.
current CPU frequency is 600 MHz.
cpufreq stats: 600 MHz:49.87%, 1000 MHz:50.13% (3)

讓我試試nbench，結果如下。
BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)

TEST                : Iterations/sec. : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :          495.76 :      12.71 :       4.18
STRING SORT         :          41.141 :      18.38 :       2.85
BITFIELD            :      1.4135e+08 :      24.25 :       5.06
FP EMULATION        :          77.569 :      37.22 :       8.59
FOURIER             :            5241 :       5.96 :       3.35
ASSIGNMENT          :          7.5314 :      28.66 :       7.43
IDEA                :          1443.8 :      22.08 :       6.56
HUFFMAN             :          725.78 :      20.13 :       6.43
NEURAL NET          :          6.8871 :      11.06 :       4.65
LU DECOMPOSITION    :          250.88 :      13.00 :       9.39
==============ORIGINAL BYTEMARK RESULTS==============
INTEGER INDEX       : 22.234
FLOATING-POINT INDEX: 9.499
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==================LINUX DATA BELOW===================
CPU                 : 4 CPU ARMv7 Processor rev 5 (v7l)
L2 Cache            :
OS                  : Linux 3.18.7-v7+
C compiler          : gcc version 4.6.3 (Debian 4.6.3-14+rpi1)
libc                : libc-2.13.so
MEMORY INDEX        : 4.749
INTEGER INDEX       : 6.235
FLOATING-POINT INDEX: 5.268
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38

然後跟舊板子
TEST                : Iterations/sec. : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :           200.4 :       5.14 :       1.69
STRING SORT         :          31.472 :      14.06 :       2.18
BITFIELD            :      8.8785e+07 :      15.23 :       3.18
FP EMULATION        :          45.509 :      21.84 :       5.04
FOURIER             :          2056.4 :       2.34 :       1.31
ASSIGNMENT          :          2.3939 :       9.11 :       2.36
IDEA                :          669.29 :      10.24 :       3.04
HUFFMAN             :          414.53 :      11.49 :       3.67
NEURAL NET          :          3.1213 :       5.01 :       2.11
LU DECOMPOSITION    :           72.68 :       3.77 :       2.72
==============ORIGINAL BYTEMARK RESULTS==============
INTEGER INDEX       : 11.448
FLOATING-POINT INDEX: 3.534
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==================LINUX DATA BELOW===================
CPU                 :
L2 Cache            :
OS                  : Linux 3.1.9+
C compiler          : gcc-4.7.real
libc                : libc-2.13.so
MEMORY INDEX        : 2.539
INTEGER INDEX       : 3.121
FLOATING-POINT INDEX: 1.960可以看見有很明顯的提昇，當然啦，相關因素很多，包括建置編譯nbench的編譯器選項，rpi板子超頻時脈，有無執行其他背景程式，等等，另外也跟測試項目有關，

另外可使用sysbench來測試有無多執行緒的效能，底下是1個執行緒的結果。
pi@raspberrypi ~ $ sysbench --test=threads run
sysbench 0.4.12: multi-threaded system evaluation benchmark

Running the test with following options:
Number of threads: 1

Doing thread subsystem performance test
Thread yields per test: 1000 Locks used: 8
Threads started!
Done.

Test execution summary:
    total time:                          18.5008s
    total number of events:              10000
    total time taken by event execution: 18.4875
    per-request statistics:
         min:                                  1.80ms
         avg:                                  1.85ms
         max:                                  9.61ms
         approx. 95 percentile:               1.89ms

Threads fairness:
    events (avg/stddev):           10000.0000/0.00
    execution time (avg/stddev):   18.4875/0.00

底下是4個執行緒的結果，
pi@raspberrypi ~ $ sysbench --num-threads=4 --test=threads run
sysbench 0.4.12: multi-threaded system evaluation benchmark

Running the test with following options:
Number of threads: 4

Doing thread subsystem performance test
Thread yields per test: 1000 Locks used: 8
Threads started!
Done.

Test execution summary:
    total time:                          4.8727s
    total number of events:              10000
    total time taken by event execution: 19.4634
    per-request statistics:
         min:                                  1.88ms
         avg:                                  1.95ms
         max:                                 10.30ms
         approx. 95 percentile:               1.96ms

Threads fairness:
    events (avg/stddev):           2500.0000/7.68
    execution time (avg/stddev):   4.8659/0.00

非常明顯，有了四核心之後，多執行緒的程式當然就跑得快。其他還有測試JavaScript的評分、測試GPU的效能表現、等等，有興趣者可自行嘗試。

最後提醒一句，跑得快也就代表比較耗電，記得提供穩定的電源與線材。

參考資料：