Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Usenet -> c.d.o.server -> Re: Sun T1 Processer - Cool Threads
Chan wrote:
> Any one using these T1 processer for Database application. If yes, then
> did you see any deferrence compare with other sparc processer. Thx.
I've done some TPC-C benchmarks with orabm tool from http://www.linxcel.co.uk/orabm. Only CPU/memory stress tests was the goal. The following list shows the configurations.
Sys A: 4xAMD 2core 875 2,2GHz (SLES9) Sys B: 4xAMD 2core 875 2,2GHz (SLES8) Sys C: Sun T1 6core 1,0 GHz (Sol10) Sys D: Sun T1 8core 1,0 GHz (Sol10) NOT tested, extrapolated from 6 core Sys E: Sun T1 8core 1,2 GHz (Sol10) NOT tested, extrapolated from 6 core Sys F: 4xSparc-IV 2core (V490, 1,35 GHz) Sol10/Ora10g) Sys G: 4xSparc-IV 2core (V490, 1,35 GHz) Sol8/Ora9i)
The test on Sys F runs with 10.1.0.4, all other tests runs with 9.2.0.7. A special test on my single core PC results in much better 10g performance on the same hardware than 9i with orabm benchmark.
The test was run as
./orabm 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,... 100000
Proc-Anz are the count of parallel Oracle-processes. The results are in TPS, generated via "sh orabm_tps.sh bm.SID.log".
The results and the feeling shows that the single performance on SUN T1 of one oracle process (one thread) are not competitive (enough), but the aggregate performance increases as more parallel CPU intensive processes are running.
I would not prefer running Oracle on Sun-T1 if performance is a theme, because of the low (peak) performance...
Proc-Anz Sys A Sys B Sys C
1 952 851 197 2 1798 1684 395 3 2688 2416 595 4 3341 3201 788 5 4133 3987 983 6 4841 4630 1174 7 5555 5394 1304 8 6164 6113 1435 9 6133 6000 1565 10 6089 5964 1688 11 6040 5882 1810 12 6081 5820 1931 13 6048 5799 2014 14 6049 5800 2092 15 6036 5765 2171 16 6018 5761 2238 17 6028 5763 2314 18 6026 5754 2397 19 6026 5770 2440 20 6024 5745 2498 21 6038 5732 2552 22 6043 5765 2588 23 6047 5771 2642 24 6046 5751 2697 25 2690 26 2690 27 2693 28 2693 29 2686 30 2687 31 2682 32 2679
2. table
Proc-Anz Sys D Sys E Sys F Sys G
1 197 236 428 383 2 395 474 804 731 3 595 714 1182 1070 4 788 946 1540 1392 5 983 1180 1870 1695 6 1174 1409 2189 1992 7 1365 1638 2507 2289 8 1556 1867 2813 2534 9 1687 2024 2817 2528 10 1818 2182 2821 2530 11 1949 2339 2815 2519 12 2080 2496 2809 2513 13 2202 2642 2815 2507 14 2324 2789 2504 15 2446 2935 2486 16 2568 3082 2496 17 2648 3178 2493 18 2728 3274 2499 19 2808 3370 2490 20 2888 3466 2498 21 2963 3556 2490 22 3038 3646 23 3113 3736 24 3188 3826 25 3253 3904 26 3318 3982 27 3383 4060 28 3448 4138 29 3500 4200 30 3552 4262 31 3604 4325 32 3656 4387
Eberhard Niendorf Received on Mon Mar 13 2006 - 18:01:56 CST
![]() |
![]() |