[BENCHMARK] 2.5.64 performance on LMbench

Aniruddha M Marathe (aniruddha.marathe@wipro.com)
Fri, 7 Mar 2003 13:01:24 +0530


Performance is not varying much since 2.5.56. Again there is no considerable performance variation for 2.5.64

L M B E N C H 2 . 0 S U M M A R Y
------------------------------------
(Alpha software, do not distribute)

Basic system parameters
----------------------------------------------------
Host OS Description Mhz

--------- ------------- ----------------------- ----
benchtest Linux 2.5.64 i686-pc-linux-gnu 790
benchtest Linux 2.5.64 i686-pc-linux-gnu 790
benchtest Linux 2.5.64 i686-pc-linux-gnu 790
benchtest Linux 2.5.64 i686-pc-linux-gnu 790
benchtest Linux 2.5.64 i686-pc-linux-gnu 790

Processor, Processes - times in microseconds - smaller is better
----------------------------------------------------------------
Host OS Mhz null null open selct sig sig fork exec sh
call I/O stat clos TCP inst hndl proc proc proc
--------- ------------- ---- ---- ---- ---- ---- ----- ---- ---- ---- ---- ----
benchtest Linux 2.5.64 790 0.44 0.83 7.44 8.89 34 1.24 3.92 276 1277 6581
benchtest Linux 2.5.64 790 0.44 0.81 7.20 8.65 35 1.26 3.92 285 1300 6621
benchtest Linux 2.5.64 790 0.44 0.84 7.39 8.89 36 1.24 3.92 314 1292 6711
benchtest Linux 2.5.64 790 0.44 0.81 7.26 8.66 35 1.24 3.92 320 1294 6608
benchtest Linux 2.5.64 790 0.44 0.81 6.99 8.70 33 1.23 3.92 299 1289 6617

Context switching - times in microseconds - smaller is better
-------------------------------------------------------------
Host OS 2p/0K 2p/16K 2p/64K 8p/16K 8p/64K 16p/16K 16p/64K
ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw
--------- ------------- ----- ------ ------ ------ ------ ------- -------
benchtest Linux 2.5.64 1.420 4.8700 16 15 176 37 177
benchtest Linux 2.5.64 1.400 4.8100 14 11 180 44 180
benchtest Linux 2.5.64 1.450 5.2400 40 19 175 41 180
benchtest Linux 2.5.64 1.440 4.9200 14 16 181 38 180
benchtest Linux 2.5.64 1.380 4.9000 14 8.9200 180 44 180

*Local* Communication latencies in microseconds - smaller is better
-------------------------------------------------------------------
Host OS 2p/0K Pipe AF UDP RPC/ TCP RPC/ TCP
ctxsw UNIX UDP TCP conn
--------- ------------- ----- ----- ---- ----- ----- ----- ----- ----
benchtest Linux 2.5.64 1.420 7.662 13 25 46 32 58 3.1M
benchtest Linux 2.5.64 1.400 8.260 13 25 46 32 58 3.1M
benchtest Linux 2.5.64 1.450 8.382 13 25 46 31 58 3.1M
benchtest Linux 2.5.64 1.440 8.355 14 25 46 32 58 3.1M
benchtest Linux 2.5.64 1.380 8.215 13 25 46 32 58 3.1M

File & VM system latencies in microseconds - smaller is better
--------------------------------------------------------------
Host OS 0K File 10K File Mmap Prot Page
Create Delete Create Delete Latency Fault Fault
--------- ------------- ------ ------ ------ ------ ------- ----- -----
benchtest Linux 2.5.64 91 31 343 82 605 0.662 4.00000
benchtest Linux 2.5.64 91 31 329 83 598 0.668 4.00000
benchtest Linux 2.5.64 91 31 331 83 610 0.661 4.00000
benchtest Linux 2.5.64 91 31 345 83 609 0.661 4.00000
benchtest Linux 2.5.64 91 31 336 84 615 0.664 4.00000

*Local* Communication bandwidths in MB/s - bigger is better
-----------------------------------------------------------
Host OS Pipe AF TCP File Mmap Bcopy Bcopy Mem Mem
UNIX reread reread (libc) (hand) read write
--------- ------------- ---- ---- ---- ------ ------ ------ ------ ---- -----
benchtest Linux 2.5.64 577 436 47 297 356 124 113 355 170
benchtest Linux 2.5.64 519 184 52 293 353 123 112 352 169
benchtest Linux 2.5.64 585 556 55 288 353 123 112 352 169
benchtest Linux 2.5.64 581 602 52 296 351 123 112 352 169
benchtest Linux 2.5.64 410 451 52 290 352 123 112 352 169

Memory latencies in nanoseconds - smaller is better
(WARNING - may not be correct, check graphs)
---------------------------------------------------
Host OS Mhz L1 $ L2 $ Main mem Guesses
--------- ------------- ---- ----- ------ -------- -------
benchtest Linux 2.5.64 790 3.798 8.8820 175
benchtest Linux 2.5.64 790 3.797 8.8820 176
benchtest Linux 2.5.64 790 3.798 55 176
benchtest Linux 2.5.64 790 3.808 8.8720 176
benchtest Linux 2.5.64 790 3.798 8.8810 176

Aniruddha Marathe
WIPRO Technologies, India
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/