Hello,
This is the summary report of four lmbench runs each. I can
provide the detailed report if there is interest. It appears
that riel is better for latency, classzone for throughput.
Machine is Athlon-700 in MS-6195, 64k L1, 512k L2, 128M RAM,
128M swap.
Caveat: HZ=100 seems pretty coarse for this speed. I don't
think there are many significant bits in some of the
timings. There are hints of aliasing in portions of the
results.
Cheers,
Tom
L M B E N C H 1 . 9 S U M M A R Y
------------------------------------
(Alpha software, do not distribute)
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 inst hndl proc proc proc
--------- ------------- ---- ---- ---- ---- ---- ----- ---- ---- ---- ---- ----
class-tex Linux 2.4.0-t 705 0.4 0.6 3 5 0.02K 1.1 2 0.2K 1K 6K
class-tex Linux 2.4.0-t 705 0.4 0.5 5 6 0.04K 1.1 2 0.4K 2K 6K
class-tex Linux 2.4.0-t 705 0.4 0.7 4 6 0.04K 1.1 2 0.4K 2K 6K
class-tex Linux 2.4.0-t 705 0.4 0.5 3 6 0.04K 1.1 2 0.2K 1K 6K
riel-text Linux 2.4.0-t 705 0.4 0.5 4 6 0.02K 1.1 2 0.2K 1K 6K
riel-text Linux 2.4.0-t 705 0.4 0.5 3 6 0.03K 1.1 2 0.3K 1K 6K
riel-text Linux 2.4.0-t 705 0.4 0.5 4 6 0.06K 1.1 2 0.3K 1K 6K
riel-text Linux 2.4.0-t 705 0.4 0.5 5 6 0.04K 1.1 2 0.2K 1K 6K
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
--------- ------------- ----- ------ ------ ------ ------ ------- -------
class-tex Linux 2.4.0-t 1 15 97 45 184 52 185
class-tex Linux 2.4.0-t 2 13 88 44 186 50 179
class-tex Linux 2.4.0-t 2 6 83 18 149 39 166
class-tex Linux 2.4.0-t 1 6 90 14 185 24 183
riel-text Linux 2.4.0-t 1 4 66 18 104 35 170
riel-text Linux 2.4.0-t 1 4 109 39 153 43 162
riel-text Linux 2.4.0-t 1 4 47 14 116 24 164
riel-text Linux 2.4.0-t 2 4 47 16 129 23 165
*Local* Communication latencies in microseconds - smaller is better
-------------------------------------------------------------------
Host OS 2p/0K Pipe AF UDP RPC/ TCP RPC/ TCP
ctxsw UNIX UDP TCP conn
--------- ------------- ----- ----- ---- ----- ----- ----- ----- ----
class-tex Linux 2.4.0-t 1 6 11 20 134 100
class-tex Linux 2.4.0-t 2 6 10 17 132 101
class-tex Linux 2.4.0-t 2 6 11 19 179 106
class-tex Linux 2.4.0-t 1 5 11 20 127 98
riel-text Linux 2.4.0-t 1 5 13 19 150 95
riel-text Linux 2.4.0-t 1 6 11 21 151 113
riel-text Linux 2.4.0-t 1 7 14 21 156 120
riel-text Linux 2.4.0-t 2 6 14 22 154 118
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
--------- ------------- ------ ------ ------ ------ ------- ----- -----
class-tex Linux 2.4.0-t 6 1 12 2 759 0 0.0K
class-tex Linux 2.4.0-t 6 1 12 2 802 0 0.0K
class-tex Linux 2.4.0-t 6 1 13 2 968 0 0.0K
class-tex Linux 2.4.0-t 6 1 12 2 721 0 0.0K
riel-text Linux 2.4.0-t 7 1 14 2 839 0 0.0K
riel-text Linux 2.4.0-t 6 1 12 2 847 0 0.0K
riel-text Linux 2.4.0-t 6 1 12 2 851 0 0.0K
riel-text Linux 2.4.0-t 7 1 13 2 960 1 0.0K
*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
--------- ------------- ---- ---- ---- ------ ------ ------ ------ ---- -----
class-tex Linux 2.4.0-t 134 76 90 118 308 192 192 308 290
class-tex Linux 2.4.0-t 378 115 110 160 308 192 192 308 290
class-tex Linux 2.4.0-t 203 96 54 145 308 192 192 308 290
class-tex Linux 2.4.0-t 128 80 52 126 308 191 192 308 290
riel-text Linux 2.4.0-t 169 107 90 163 308 194 193 308 290
riel-text Linux 2.4.0-t 100 73 68 120 308 194 193 308 290
riel-text Linux 2.4.0-t 371 108 94 187 308 194 194 308 289
riel-text Linux 2.4.0-t 195 86 77 163 308 194 194 308 289
Memory latencies in nanoseconds - smaller is better
(WARNING - may not be correct, check graphs)
---------------------------------------------------
Host OS Mhz L1 $ L2 $ Main mem Guesses
--------- ------------- --- ---- ---- -------- -------
class-tex Linux 2.4.0-t 705 4 175 218 No L2 cache?
class-tex Linux 2.4.0-t 705 4 171 221
class-tex Linux 2.4.0-t 705 4 171 220
class-tex Linux 2.4.0-t 705 4 162 219
riel-text Linux 2.4.0-t 705 4 113 222
riel-text Linux 2.4.0-t 705 4 141 222
riel-text Linux 2.4.0-t 705 4 32 222
riel-text Linux 2.4.0-t 705 4 45 222
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Jun 23 2000 - 21:00:22 EST