知り合いに型落ちのHP Proliant DL380 G7の構築を依頼されたため、構築しつつベンチを取ってみた。
Xeon E5620 2.4GHz 4core/8thread x 2
DDR3-1066 4GB x 4
SAS 600GB 10krpm x 8 RAID5 @ P420
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 |
------------------------------------------------------------------------ 16 CPUs in system; running 1 parallel copy of tests Dhrystone 2 using register variables 23086264.4 lps (10.0 s, 7 samples) Double-Precision Whetstone 2867.0 MWIPS (10.0 s, 7 samples) Execl Throughput 2607.9 lps (29.9 s, 2 samples) File Copy 1024 bufsize 2000 maxblocks 728467.0 KBps (30.0 s, 2 samples) File Copy 256 bufsize 500 maxblocks 194072.5 KBps (30.0 s, 2 samples) File Copy 4096 bufsize 8000 maxblocks 1637842.7 KBps (30.0 s, 2 samples) Pipe Throughput 1366677.0 lps (10.0 s, 7 samples) Pipe-based Context Switching 145368.0 lps (10.0 s, 7 samples) Process Creation 4446.9 lps (30.0 s, 2 samples) Shell Scripts (1 concurrent) 4894.4 lpm (60.0 s, 2 samples) Shell Scripts (8 concurrent) 2744.2 lpm (60.0 s, 2 samples) System Call Overhead 1611113.1 lps (10.0 s, 7 samples) System Benchmarks Index Values BASELINE RESULT INDEX Dhrystone 2 using register variables 116700.0 23086264.4 1978.3 Double-Precision Whetstone 55.0 2867.0 521.3 Execl Throughput 43.0 2607.9 606.5 File Copy 1024 bufsize 2000 maxblocks 3960.0 728467.0 1839.6 File Copy 256 bufsize 500 maxblocks 1655.0 194072.5 1172.6 File Copy 4096 bufsize 8000 maxblocks 5800.0 1637842.7 2823.9 Pipe Throughput 12440.0 1366677.0 1098.6 Pipe-based Context Switching 4000.0 145368.0 363.4 Process Creation 126.0 4446.9 352.9 Shell Scripts (1 concurrent) 42.4 4894.4 1154.3 Shell Scripts (8 concurrent) 6.0 2744.2 4573.6 System Call Overhead 15000.0 1611113.1 1074.1 ======== System Benchmarks Index Score 1097.2 ------------------------------------------------------------------------ 16 CPUs in system; running 16 parallel copies of tests Dhrystone 2 using register variables 198252404.5 lps (10.0 s, 7 samples) Double-Precision Whetstone 40272.6 MWIPS (9.6 s, 7 samples) Execl Throughput 29858.4 lps (29.9 s, 2 samples) File Copy 1024 bufsize 2000 maxblocks 535705.5 KBps (30.0 s, 2 samples) File Copy 256 bufsize 500 maxblocks 146569.7 KBps (30.0 s, 2 samples) File Copy 4096 bufsize 8000 maxblocks 1703923.5 KBps (30.0 s, 2 samples) Pipe Throughput 11826342.8 lps (10.0 s, 7 samples) Pipe-based Context Switching 2335343.3 lps (10.0 s, 7 samples) Process Creation 64995.4 lps (30.0 s, 2 samples) Shell Scripts (1 concurrent) 39268.1 lpm (60.0 s, 2 samples) Shell Scripts (8 concurrent) 5510.5 lpm (60.1 s, 2 samples) System Call Overhead 6557937.7 lps (10.0 s, 7 samples) System Benchmarks Index Values BASELINE RESULT INDEX Dhrystone 2 using register variables 116700.0 198252404.5 16988.2 Double-Precision Whetstone 55.0 40272.6 7322.3 Execl Throughput 43.0 29858.4 6943.8 File Copy 1024 bufsize 2000 maxblocks 3960.0 535705.5 1352.8 File Copy 256 bufsize 500 maxblocks 1655.0 146569.7 885.6 File Copy 4096 bufsize 8000 maxblocks 5800.0 1703923.5 2937.8 Pipe Throughput 12440.0 11826342.8 9506.7 Pipe-based Context Switching 4000.0 2335343.3 5838.4 Process Creation 126.0 64995.4 5158.4 Shell Scripts (1 concurrent) 42.4 39268.1 9261.3 Shell Scripts (8 concurrent) 6.0 5510.5 9184.1 System Call Overhead 15000.0 6557937.7 4372.0 ======== System Benchmarks Index Score 5118.9 |
シングルスレッドの演算性能は先日取ったE3-1290v2の半分、システム全体のパラレル演算能力は1割ほど高い程度の結果。
消費電力あたりの性能を考えると素直にE3を乗せたの方が良好と言えるが、最大メモリ搭載量や拡張スロット、ドライブベイ等の点を考慮すると、まぁありな選択肢ではあるかと思う。
Xeon56xx系でも8Cのデュアルとかになれば現行E3の倍以上の演算能力になるわけで、まぁその辺のコンフィギュレーションになればまだまだ実戦投入可能かな？
(290)
