# vim /boot/config.txt #uncomment to overclock the arm. 700 MHz is the default. arm_freq=1000 #CPUの周波数 force_turbo=1 #CPUの周波数を動的にしない & 最小CPU周波数設定より下にしないようにする 有効1 # for more options see http://elinux.org/RPi_config.txt core_freq=500 #GPUの周波数。ARMのL2キャッシュに影響がある？（It has an impact on ARM performance since it drives L2 cache.） sdram_freq=600 #SDRAMの周波数 over_voltage=8 #ARM/GPUの電圧で0で1.2V、6より上はforce_turboか、current_limit_overrideを明示したら有効になる
Raspberry Piってオーバークロック出来るらしいので「Unix Bench」を使って比較してみた。
オーバークロックするには /boot/config.txt を編集する。
参考 : http://elinux.org/RPiconfig
pidora にUnix Benchを入れる。
[source]
apt-get install libx11-dev libgl1-mesa-dev libxext-dev perl perl-modules
wget http://byte-unixbench.googlecode.com/files/unixbench-5.1.2.tar.gz
tar xvf unixbench-5.1.2.tar.gz
cd unixbench-5.1.2
./Run #スタート！
[/source]
最初は 700Mhz の設定から
[source]
1 CPU in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 1680000.0 lps (10.0 s, 7 samples)
Double-Precision Whetstone 269.8 MWIPS (9.9 s, 7 samples)
Execl Throughput 319.3 lps (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 51317.6 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 15562.4 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 127048.4 KBps (30.0 s, 2 samples)
Pipe Throughput 171043.8 lps (10.0 s, 7 samples)
Pipe-based Context Switching 27494.0 lps (10.0 s, 7 samples)
Process Creation 1004.7 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 546.7 lpm (60.1 s, 2 samples)
Shell Scripts (8 concurrent) 69.7 lpm (60.3 s, 2 samples)
System Call Overhead 319707.1 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 1680000.0 144.0
Double-Precision Whetstone 55.0 269.8 49.1
Execl Throughput 43.0 319.3 74.3
File Copy 1024 bufsize 2000 maxblocks 3960.0 51317.6 129.6
File Copy 256 bufsize 500 maxblocks 1655.0 15562.4 94.0
File Copy 4096 bufsize 8000 maxblocks 5800.0 127048.4 219.0
Pipe Throughput 12440.0 171043.8 137.5
Pipe-based Context Switching 4000.0 27494.0 68.7
Process Creation 126.0 1004.7 79.7
Shell Scripts (1 concurrent) 42.4 546.7 128.9
Shell Scripts (8 concurrent) 6.0 69.7 116.1
System Call Overhead 15000.0 319707.1 213.1
System Benchmarks Index Score 110.6
[/source]
次に 1000Mhz の設定
[source]
1 CPU in system; running 1 parallel copy of tests
Dhrystone 2 using register variables 2409318.7 lps (10.0 s, 7 samples)
Double-Precision Whetstone 385.5 MWIPS (9.9 s, 7 samples)
Execl Throughput 380.5 lps (29.9 s, 2 samples)
File Copy 1024 bufsize 2000 maxblocks 65480.4 KBps (30.0 s, 2 samples)
File Copy 256 bufsize 500 maxblocks 20513.2 KBps (30.0 s, 2 samples)
File Copy 4096 bufsize 8000 maxblocks 156624.6 KBps (30.0 s, 2 samples)
Pipe Throughput 245173.0 lps (10.0 s, 7 samples)
Pipe-based Context Switching 37351.6 lps (10.0 s, 7 samples)
Process Creation 1196.1 lps (30.0 s, 2 samples)
Shell Scripts (1 concurrent) 676.8 lpm (60.1 s, 2 samples)
Shell Scripts (8 concurrent) 86.4 lpm (60.4 s, 2 samples)
System Call Overhead 571589.6 lps (10.0 s, 7 samples)
System Benchmarks Index Values BASELINE RESULT INDEX
Dhrystone 2 using register variables 116700.0 2409318.7 206.5
Double-Precision Whetstone 55.0 385.5 70.1
Execl Throughput 43.0 380.5 88.5
File Copy 1024 bufsize 2000 maxblocks 3960.0 65480.4 165.4
File Copy 256 bufsize 500 maxblocks 1655.0 20513.2 123.9
File Copy 4096 bufsize 8000 maxblocks 5800.0 156624.6 270.0
Pipe Throughput 12440.0 245173.0 197.1
Pipe-based Context Switching 4000.0 37351.6 93.4
Process Creation 126.0 1196.1 94.9
Shell Scripts (1 concurrent) 42.4 676.8 159.6
Shell Scripts (8 concurrent) 6.0 86.4 144.1
System Call Overhead 15000.0 571589.6 381.1
System Benchmarks Index Score 147.8
[/source]
効果は若干あるう～
luispc.asia の方は 1000Mhz で動かしております。
ラズベリーパイ購入先：RSコンポーネンツ