Cross-Platform Benchmarks

Blue-Fox

Die 68030 CPU hat eine MMU. Sogar eine FPU hab ich auf der Turbokarte. Eine 688828 oder so.

**oreissig**

Zitat von mrshadowtux

Schade, dass es den Studio-Compiler nicht für Sparc Linux gibt.

hmm, vielleicht kann der unter Linux/x86 trotzdem SPARC-Code erzeugen?

Zitat von Blue-Fox

Die 68030 CPU hat eine MMU. Sogar eine FPU hab ich auf der Turbokarte. Eine 688828 oder so.

wenn du eine CPU-Karte mit >=030 oder eine 68851 PMMU hast sollte es gehen, beides ist halt bei den verbreitetsten amigas standardmäßig nich der Fall

Blue-Fox

Ich habe eine A2630 in meiner Amiga 2000 verbaut.
Die hat eine 68030 CPU mit einer PMMU intern verbaut, und eine 68882 FPU.
Außerdem sind 1MB Chipram und 8MB Fastram verbaut.
Für AMIX reicht das, aber für nBench?
Wenn ich sie mal enger am Wickel habe, werd ich mal schauen, was ich tun kann.

Anbei sind die Ergebnisse von openSSL, cat /proc/cpuinfo und uname -a.
openSSL:

Spoiler anzeigen

Code

OpenSSL 1.0.0e 6 Sep 2011
built on: Sat Nov 19 11:13:15 CET 2011
options:bn(64,32) md2(int) rc4(ptr,char) des(idx,risc1,16,long) aes(partial) idea(int) blowfish(idx) 
compiler: gcc -fPIC -DOPENSSL_PIC -DZLIB -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -Wa,--noexecstack -mtune=powerpc -mpowerpc-gfxopt -O2 -pipe -DB_ENDIAN -DTERMIO -O3 -Wall -DSHA1_ASM -DSHA256_ASM -DAES_ASM
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
md2               1097.78k     2300.92k     3152.98k     3478.87k     3588.10k
mdc2              3541.54k     4176.38k     4371.71k     4423.34k     4440.06k
md4              10846.80k    37711.59k   105789.35k   192930.47k   254186.84k
md5               9053.36k    30721.47k    83914.15k   147932.16k   190237.35k
hmac(md5)         8676.74k    29836.31k    82201.51k   146683.56k   190084.44k
sha1             10286.39k    34562.47k    92914.43k   161245.87k   205037.57k
rmd160            8417.89k    25944.17k    61152.51k    92594.86k   109005.48k
rc4              84340.28k    95919.64k    99321.00k   100219.56k   100477.61k
des cbc          21052.93k    22404.89k    22783.06k    22877.87k    22907.56k
des ede3          8315.08k     8541.25k     8600.58k     8616.62k     8620.71k
idea cbc         29990.94k    32775.70k    33573.63k    33782.78k    33843.88k
seed cbc         27384.38k    29311.49k    29834.33k    29927.42k    29996.37k
rc2 cbc          17125.09k    18006.89k    18323.95k    18278.74k    18279.08k
rc5-32/12 cbc        0.00         0.00         0.00         0.00         0.00 
blowfish cbc     38188.67k    42815.77k    44190.38k    44548.78k    44649.13k
cast cbc         34719.21k    38505.51k    39614.46k    39903.23k    39987.88k
aes-128 cbc      30735.43k    33511.75k    34264.66k    34447.36k    34542.93k
aes-192 cbc      26721.41k    28745.05k    29391.02k    29528.06k    29600.43k
aes-256 cbc      23693.40k    25245.91k    25737.22k    25839.96k    25894.91k
camellia-128 cbc    34772.22k    38439.38k    39501.65k    39703.89k    39834.97k
camellia-192 cbc    28899.26k    31390.70k    32095.83k    32228.35k    32317.44k
camellia-256 cbc    28898.12k    31390.91k    32095.06k    32228.69k    32317.44k
sha256           13195.18k    32417.56k    60541.53k    77652.99k    84631.55k
sha512            1279.65k     5130.13k     7099.99k     9577.47k    10660.52k
whirlpool          778.81k     1597.48k     2558.21k     3019.78k     3254.95k
aes-128 ige      29588.04k    32899.58k    33865.90k    34075.99k    34177.02k
aes-192 ige      25886.51k    28382.72k    29102.42k    29253.29k    29332.82k
aes-256 ige      23003.42k    25040.61k    25508.27k    25628.33k    25687.38k
                  sign    verify    sign/s verify/s
rsa  512 bits 0.001197s 0.000103s    835.5   9705.6
rsa 1024 bits 0.006246s 0.000310s    160.1   3222.6
rsa 2048 bits 0.038206s 0.001057s     26.2    945.7
rsa 4096 bits 0.256250s 0.003798s      3.9    263.3
                  sign    verify    sign/s verify/s
dsa  512 bits 0.001042s 0.001177s    959.5    849.5
dsa 1024 bits 0.003085s 0.003632s    324.1    275.3
dsa 2048 bits 0.010442s 0.012541s     95.8     79.7
                              sign    verify    sign/s verify/s
 160 bit ecdsa (secp160r1)   0.0006s   0.0030s   1552.5    336.9
 192 bit ecdsa (nistp192)   0.0007s   0.0032s   1476.2    314.8
 224 bit ecdsa (nistp224)   0.0009s   0.0042s   1140.8    239.5
 256 bit ecdsa (nistp256)   0.0011s   0.0055s    904.8    182.9
 384 bit ecdsa (nistp384)   0.0025s   0.0132s    407.1     75.6
 521 bit ecdsa (nistp521)   0.0052s   0.0279s    191.3     35.8
 163 bit ecdsa (nistk163)   0.0027s   0.0060s    367.9    166.3
 233 bit ecdsa (nistk233)   0.0054s   0.0112s    183.7     89.4
 283 bit ecdsa (nistk283)   0.0083s   0.0207s    119.9     48.3
 409 bit ecdsa (nistk409)   0.0195s   0.0473s     51.2     21.1
 571 bit ecdsa (nistk571)   0.0448s   0.1100s     22.3      9.1
 163 bit ecdsa (nistb163)   0.0027s   0.0064s    370.8    157.1
 233 bit ecdsa (nistb233)   0.0054s   0.0123s    185.8     81.4
 283 bit ecdsa (nistb283)   0.0083s   0.0230s    120.0     43.4
 409 bit ecdsa (nistb409)   0.0196s   0.0537s     51.0     18.6
 571 bit ecdsa (nistb571)   0.0447s   0.1255s     22.4      8.0
                              op      op/s
 160 bit ecdh (secp160r1)   0.0026s    390.5
 192 bit ecdh (nistp192)   0.0027s    376.2
 224 bit ecdh (nistp224)   0.0035s    285.6
 256 bit ecdh (nistp256)   0.0046s    216.0
 384 bit ecdh (nistp384)   0.0111s     89.8
 521 bit ecdh (nistp521)   0.0232s     43.2
 163 bit ecdh (nistk163)   0.0029s    344.5
 233 bit ecdh (nistk233)   0.0055s    181.7
 283 bit ecdh (nistk283)   0.0101s     99.4
 409 bit ecdh (nistk409)   0.0233s     42.9
 571 bit ecdh (nistk571)   0.0546s     18.3
 163 bit ecdh (nistb163)   0.0032s    317.1
 233 bit ecdh (nistb233)   0.0060s    165.4
 283 bit ecdh (nistb283)   0.0115s     86.6
 409 bit ecdh (nistb409)   0.0267s     37.5
 571 bit ecdh (nistb571)   0.0630s     15.9

Alles anzeigen

/proc/cpuinfo:

Spoiler anzeigen

processor : 0
cpu : 7447A, altivec supported
clock : 1249.999995MHz
revision : 1.1 (pvr 8003 0101)
bogomips : 83.24

total bogomips : 83.24
timebase : 41620997
platform : PowerMac
model : PowerMac6,4
machine : PowerMac6,4
motherboard : PowerMac6,4 MacRISC3 Power Macintosh
detected as : 287 (eMac)
pmac flags : 00000010
L2 cache : 512K unified
pmac-generation : NewWorld
Memory : 1280 MB

uname -a:
Linux Rhea 3.1.4-1-ARCH #1 SMP PREEMPT Wed Nov 30 17:27:25 CET 2011 ppc 7447A, altivec supported PowerMac6,4 GNU/Linux

**oreissig**

meine c8000 unter HP-UX 11.31 mit gcc

Code

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)


TEST                : Iterations/sec.  : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :          488.28  :      12.52  :       4.11
STRING SORT         :          155.96  :      69.69  :      10.79
BITFIELD            :       1.078e+08  :      18.49  :       3.86
FP EMULATION        :          107.24  :      51.46  :      11.87
FOURIER             :           19781  :      22.50  :      12.64
ASSIGNMENT          :          11.862  :      45.14  :      11.71
IDEA                :          735.88  :      11.26  :       3.34
HUFFMAN             :          804.24  :      22.30  :       7.12
NEURAL NET          :          18.156  :      29.17  :      12.27
LU DECOMPOSITION    :          583.68  :      30.24  :      21.83
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 26.594
FLOATING-POINT INDEX: 27.071
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : 900MHz PA-8800 (Dualcore)
L2 Cache            : 32MB
OS                  : HP-UX B.11.31
C compiler          : gcc version 4.6.0 (GCC)
libc                :
MEMORY INDEX        : 7.872
INTEGER INDEX       : 5.839
FLOATING-POINT INDEX: 15.014
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

Alles anzeigen

optimiert auf PA-8000, denn der aktuellste gcc kann leider den code auf PA-RISC für nichts neueres optimieren als eine CPU, die im Jahr vor dem Pentium II rauskam

EDIT: und mal noch das in die Jahre gekommene, mitgelieferte OpenSSL
hab das gefühl, dass ich mit dem gcc da nichts hinbekommen würde, was schneller wäre

Spoiler anzeigen

Code

OpenSSL 0.9.8d 28 Sep 2006
built on: Wed Oct 11 15:32:16 IST 2006
options:bn(32,32) md2(int) rc4(ptr,int) des(ptr,risc1,16,long) aes(partial) blow                                                                                             fish(idx)
compiler: cc +Z -DOPENSSL_PIC -DZLIB -DOPENSSL_THREADS -DDSO_DL -D_REENTRANT +DA                                                                                             2.0 -Wl,+nodefaultrpath -DB_ENDIAN -DBN_DIV2W -DMD32_XARRAY -Ae +ESlit +O1 -z
available timing options: TIMES TIMEB HZ=100 [sysconf value]
timing function used: times
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
md2                659.32k     1430.86k     2020.61k     2256.21k     2329.26k
mdc2                 0.00         0.00         0.00         0.00         0.00
md4               5311.48k    18224.78k    49565.85k    87623.00k   113082.37k
md5               4247.16k    13577.83k    39054.71k    68348.59k    87790.93k
hmac(md5)         5146.41k    16501.46k    43636.31k    71257.77k    88062.64k
sha1              4751.25k    15903.87k    41400.06k    69983.43k    88202.36k
rmd160            3692.15k    10108.70k    20877.48k    28439.89k    31706.58k
rc4              69413.20k    80077.80k    83042.49k    84162.22k    84273.83k
des cbc           6449.79k     6692.86k     6734.42k     6771.03k     6777.51k
des ede3          2365.57k     2393.05k     2410.33k     2403.85k     2413.91k
idea cbc             0.00         0.00         0.00         0.00         0.00
rc2 cbc          10632.26k    11284.89k    11491.83k    11508.39k    11537.29k
rc5-32/12 cbc        0.00         0.00         0.00         0.00         0.00
blowfish cbc     23162.15k    26646.69k    27778.13k    28055.21k    28136.79k
cast cbc         12542.81k    13497.88k    13736.53k    13809.32k    13808.98k
aes-128 cbc      18217.17k    18898.05k    19163.14k    19166.83k    19251.20k
aes-192 cbc      16579.43k    17361.43k    17498.28k    17475.36k    17519.96k
aes-256 cbc      15332.67k    15925.65k    16143.27k    16190.12k    16149.94k
camellia-128 cbc        0.00         0.00         0.00         0.00         0.00                                                                                             
camellia-192 cbc        0.00         0.00         0.00         0.00         0.00                                                                                             
camellia-256 cbc        0.00         0.00         0.00         0.00         0.00                                                                                             
sha256            3361.39k     8774.81k    17150.21k    22532.78k    24733.85k
sha512            1505.54k     6022.70k    10335.94k    15353.51k    17861.29k
                  sign    verify    sign/s verify/s
rsa  512 bits 0.032362s 0.003018s     30.9    331.4
rsa 1024 bits 0.205102s 0.011002s      4.9     90.9
rsa 2048 bits 1.397500s 0.041240s      0.7     24.2
rsa 4096 bits 9.940000s 0.154462s      0.1      6.5
                  sign    verify    sign/s verify/s
dsa  512 bits 0.029910s 0.036101s     33.4     27.7
dsa 1024 bits 0.109022s 0.132895s      9.2      7.5
dsa 2048 bits 0.399600s 0.483333s      2.5      2.1

Alles anzeigen

EDIT2: okay, doch ne ganze stange fixer

Spoiler anzeigen

Code

OpenSSL 1.0.0g 18 Jan 2012
built on: Thu Feb  9 06:21:05 MEZ 2012
options:bn(64,64) rc4(idx,char) des(idx,risc1,16,int) aes(partial) idea(int) blo                                                                                             wfish(idx)
compiler: gcc -DOPENSSL_THREADS  -DDSO_DLFCN -DHAVE_DLFCN_H -march=2.0 -mschedul                                                                                             e=8000 -D_REENTRANT -Wa,--noexecstack -O3 -DB_ENDIAN -D_REENTRANT
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
md2                  0.00         0.00         0.00         0.00         0.00
mdc2              1624.31k     2409.88k     2742.02k     2837.50k     2867.20k
md4               3225.47k    12264.41k    42637.48k   111718.40k   212065.95k
md5               2683.18k    10040.46k    32737.04k    76887.04k   125389.31k
hmac(md5)         6770.04k    22957.06k    60793.17k   104889.34k   132969.81k
sha1              2920.22k    10352.63k    31215.79k    63630.68k    91340.80k
rmd160            2772.33k     9511.40k    26386.69k    47496.87k    61964.29k
rc4              87594.76k    97421.35k   100033.57k   100934.31k   101199.97k
des cbc          15435.91k    16395.90k    16588.12k    16704.85k    16733.53k
des ede3          6356.00k     6552.77k     6612.57k     6600.36k     6618.92k
idea cbc          6067.67k     6236.89k     6271.40k     6290.09k     6287.84k
seed cbc         19078.03k    21213.77k    21898.41k    22078.46k    22121.13k
rc2 cbc          10587.66k    11025.42k    11041.88k    11140.44k    11091.97k
rc5-32/12 cbc        0.00         0.00         0.00         0.00         0.00
blowfish cbc     26374.17k    29816.49k    30916.44k    31172.41k    31094.10k
cast cbc         25108.70k    28305.77k    29153.21k    29512.02k    29540.52k
aes-128 cbc      32443.26k    39504.38k    41758.98k    42219.32k    42535.59k
aes-192 cbc      29463.58k    35297.13k    36937.40k    37521.41k    37716.19k
aes-256 cbc      27074.80k    31871.02k    33298.60k    33675.26k    33690.62k
camellia-128 cbc    24774.02k    28946.65k    30084.27k    30416.55k    30411.10                                                                                             k
camellia-192 cbc    20798.06k    23365.29k    24070.97k    24371.20k    24346.62                                                                                             k
camellia-256 cbc    20586.78k    23300.59k    24112.81k    24357.50k    24359.56                                                                                             k
sha256            5717.59k    14538.11k    27526.29k    35631.79k    38907.89k
sha512            3402.15k    13675.75k    24571.56k    37070.16k    43737.09k
whirlpool         4133.59k     8912.46k    14951.68k    17923.41k    19111.94k
aes-128 ige      27840.52k    37403.31k    40988.75k    42135.03k    42308.15k
aes-192 ige      25607.61k    33564.71k    36486.23k    37240.49k    37568.51k
aes-256 ige      23837.81k    30445.91k    32867.67k    33394.65k    33776.88k
                  sign    verify    sign/s verify/s
rsa  512 bits 0.001170s 0.000089s    854.4  11243.8
rsa 1024 bits 0.005038s 0.000236s    198.5   4235.6
rsa 2048 bits 0.027370s 0.000714s     36.5   1400.8
rsa 4096 bits 0.167500s 0.002383s      6.0    419.7
                  sign    verify    sign/s verify/s
dsa  512 bits 0.000958s 0.001015s   1043.7    985.4
dsa 1024 bits 0.002357s 0.002749s    424.2    363.8
dsa 2048 bits 0.007053s 0.008443s    141.8    118.4
                              sign    verify    sign/s verify/s
 160 bit ecdsa (secp160r1)   0.0007s   0.0032s   1385.7    309.3
 192 bit ecdsa (nistp192)   0.0007s   0.0028s   1527.6    353.9
 224 bit ecdsa (nistp224)   0.0009s   0.0039s   1139.6    256.3
 256 bit ecdsa (nistp256)   0.0011s   0.0052s    906.3    191.8
 384 bit ecdsa (nistp384)   0.0020s   0.0105s    491.2     95.5
 521 bit ecdsa (nistp521)   0.0032s   0.0166s    315.9     60.2
 163 bit ecdsa (nistk163)   0.0026s   0.0052s    385.4    190.9
 233 bit ecdsa (nistk233)   0.0050s   0.0070s    201.1    141.9
 283 bit ecdsa (nistk283)   0.0077s   0.0153s    129.9     65.3
 409 bit ecdsa (nistk409)   0.0169s   0.0328s     59.3     30.5
 571 bit ecdsa (nistk571)   0.0360s   0.0721s     27.7     13.9
 163 bit ecdsa (nistb163)   0.0026s   0.0056s    388.3    178.4
 233 bit ecdsa (nistb233)   0.0049s   0.0075s    202.9    133.0
 283 bit ecdsa (nistb283)   0.0077s   0.0171s    130.6     58.5
 409 bit ecdsa (nistb409)   0.0169s   0.0366s     59.2     27.3
 571 bit ecdsa (nistb571)   0.0359s   0.0813s     27.8     12.3
                              op      op/s
 160 bit ecdh (secp160r1)   0.0027s    366.5
 192 bit ecdh (nistp192)   0.0024s    423.8
 224 bit ecdh (nistp224)   0.0033s    304.9
 256 bit ecdh (nistp256)   0.0043s    234.0
 384 bit ecdh (nistp384)   0.0088s    114.0
 521 bit ecdh (nistp521)   0.0138s     72.4
 163 bit ecdh (nistk163)   0.0025s    394.7
 233 bit ecdh (nistk233)   0.0035s    286.0
 283 bit ecdh (nistk283)   0.0076s    131.8
 409 bit ecdh (nistk409)   0.0161s     62.1
 571 bit ecdh (nistk571)   0.0357s     28.0
 163 bit ecdh (nistb163)   0.0028s    362.8
 233 bit ecdh (nistb233)   0.0037s    272.2
 283 bit ecdh (nistb283)   0.0084s    119.5
 409 bit ecdh (nistb409)   0.0182s     55.0
 571 bit ecdh (nistb571)   0.0404s     24.8

Alles anzeigen

**oreissig**

mein HP TouchPad:

Code

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)


TEST                : Iterations/sec.  : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :          381.48  :       9.78  :       3.21
STRING SORT         :          52.895  :      23.64  :       3.66
BITFIELD            :      1.2343e+08  :      21.17  :       4.42
FP EMULATION        :          45.202  :      21.69  :       5.01
FOURIER             :          3767.6  :       4.28  :       2.41
ASSIGNMENT          :          6.2839  :      23.91  :       6.20
IDEA                :            1351  :      20.66  :       6.13
HUFFMAN             :          590.82  :      16.38  :       5.23
NEURAL NET          :         0.92136  :       1.48  :       0.62
LU DECOMPOSITION    :          22.836  :       1.18  :       0.85
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 18.894
FLOATING-POINT INDEX: 1.958
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : ARMv7 Processor rev 2 (v7l)
L2 Cache            : 0
OS                  : Linux version 2.6.35-palm-tenderloin (erik@myhost) (gcc version 4.4.1 (Sourcery G++ Lite 2010q1-188) ) #12 SMP PREEMPT Tue Feb 21 00:04:17 CST 2012
C compiler          : arm-eabi-gcc (GCC) 4.4.0
libc                : Android Bionic libc
MEMORY INDEX        : 4.647
INTEGER INDEX       : 4.766
FLOATING-POINT INDEX: 1.086
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

Alles anzeigen

**Blue**

android uf touchpad

**oreissig**

achso ist btw CyanogenMod 9 Alpha 2

**Xaar**

Ich hab mal den nBench über meine rx2600 mit diversen CPU-Konfigurationen unter Debian 6.04 drübergejagt
Compiler, CFLAGS, OS-Version:

Spoiler anzeigen

gcc version 4.4.5 (Debian 4.4.5-8)

CFLAGS = -s -static -Wall -O3

Linux HP-RX2600 2.6.32-5-mckinley #1 SMP Sun May 6 08:36:37 UTC 2012 ia64 GNU/Linux

Die CFLAGS sind die, die standardmäßig eingestellt sind. Andere hab ich mangels Wissens, was es gibt (noch?) net getestet.

Hier die Ergebnisse, hinterher gleich die cpuinfo aus /proc/cpuinfo:

Itanium 2 1000MHz 1MB L3 Qualification Sample (Single)

Spoiler anzeigen

Code

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)


TEST                : Iterations/sec.  : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :           825.2  :      21.16  :       6.95
STRING SORT         :          18.541  :       8.28  :       1.28
BITFIELD            :      1.4262e+08  :      24.46  :       5.11
FP EMULATION        :          104.25  :      50.02  :      11.54
FOURIER             :           37495  :      42.64  :      23.95
ASSIGNMENT          :          10.367  :      39.45  :      10.23
IDEA                :          989.71  :      15.14  :       4.49
HUFFMAN             :           895.7  :      24.84  :       7.93
NEURAL NET          :          9.3835  :      15.07  :       6.34
LU DECOMPOSITION    :          326.72  :      16.93  :      12.22
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 22.779
FLOATING-POINT INDEX: 22.158
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : Madison 1000MHz
L2 Cache            :
OS                  : Linux 2.6.32-5-mckinley
C compiler          :
libc                :
MEMORY INDEX        : 4.062
INTEGER INDEX       : 7.313
FLOATING-POINT INDEX: 12.289
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

Alles anzeigen

Code

processor  : 0
vendor     : GenuineIntel
arch       : IA-64
family     : 31
model      : 1
model name : Madison
revision   : 5
archrev    : 0
features   : branchlong
cpu number : 0
cpu regs   : 4
cpu MHz    : 1000.000
itc MHz    : 1000.000000
BogoMIPS   : 1495.04
siblings   : 1
physical id: 0

Alles anzeigen

Itanium 2 1300MHz 3MB L3 (Dual)

Spoiler anzeigen

Code

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)


TEST                : Iterations/sec.  : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :          957.24  :      24.55  :       8.06
STRING SORT         :          24.092  :      10.77  :       1.67
BITFIELD            :      1.8535e+08  :      31.79  :       6.64
FP EMULATION        :          135.66  :      65.10  :      15.02
FOURIER             :           48761  :      55.46  :      31.15
ASSIGNMENT          :          13.475  :      51.27  :      13.30
IDEA                :            1287  :      19.68  :       5.84
HUFFMAN             :          1164.5  :      32.29  :      10.31
NEURAL NET          :           12.21  :      19.62  :       8.25
LU DECOMPOSITION    :          424.88  :      22.01  :      15.89
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 29.137
FLOATING-POINT INDEX: 28.821
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : Dual Madison 1300MHz
L2 Cache            :
OS                  : Linux 2.6.32-5-mckinley
C compiler          :
libc                :
MEMORY INDEX        : 5.280
INTEGER INDEX       : 9.243
FLOATING-POINT INDEX: 15.985
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

Alles anzeigen

Code

processor  : 0
vendor     : GenuineIntel
arch       : IA-64
family     : 31
model      : 1
model name : Madison
revision   : 5
archrev    : 0
features   : branchlong
cpu number : 0
cpu regs   : 4
cpu MHz    : 1300.000
itc MHz    : 1300.000000
BogoMIPS   : 1945.60
siblings   : 1
physical id: 0


processor  : 1
vendor     : GenuineIntel
arch       : IA-64
family     : 31
model      : 1
model name : Madison
revision   : 5
archrev    : 0
features   : branchlong
cpu number : 0
cpu regs   : 4
cpu MHz    : 1300.000
itc MHz    : 1300.000000
BogoMIPS   : 1945.60
siblings   : 1
physical id: 1

Alles anzeigen

Itanium 2 1400MHz 1.5MB L3 Qualification Sample (Single)

Spoiler anzeigen

Code

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)


TEST                : Iterations/sec.  : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :          1030.8  :      26.44  :       8.68
STRING SORT         :          25.957  :      11.60  :       1.80
BITFIELD            :      1.9955e+08  :      34.23  :       7.15
FP EMULATION        :             146  :      70.06  :      16.17
FOURIER             :           52512  :      59.72  :      33.54
ASSIGNMENT          :          14.521  :      55.26  :      14.33
IDEA                :            1386  :      21.20  :       6.29
HUFFMAN             :          1254.8  :      34.80  :      11.11
NEURAL NET          :          13.143  :      21.11  :       8.88
LU DECOMPOSITION    :          458.17  :      23.74  :      17.14
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 31.381
FLOATING-POINT INDEX: 31.047
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : Madison 1400MHz
L2 Cache            :
OS                  : Linux 2.6.32-5-mckinley
C compiler          :
libc                :
MEMORY INDEX        : 5.687
INTEGER INDEX       : 9.954
FLOATING-POINT INDEX: 17.220
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

Alles anzeigen

Code

processor  : 0
vendor     : GenuineIntel
arch       : IA-64
family     : 31
model      : 1
model name : Madison
revision   : 5
archrev    : 0
features   : branchlong
cpu number : 0
cpu regs   : 4
cpu MHz    : 1400.000
itc MHz    : 1400.000000
BogoMIPS   : 2097.15
siblings   : 1
physical id: 0

Alles anzeigen

Itanium 2 1400MHz 4MB L3 (Single)

Spoiler anzeigen

Code

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)


TEST                : Iterations/sec.  : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :          1031.3  :      26.45  :       8.69
STRING SORT         :          25.967  :      11.60  :       1.80
BITFIELD            :       1.997e+08  :      34.26  :       7.16
FP EMULATION        :          146.02  :      70.07  :      16.17
FOURIER             :           52544  :      59.76  :      33.56
ASSIGNMENT          :          14.521  :      55.26  :      14.33
IDEA                :            1386  :      21.20  :       6.29
HUFFMAN             :          1254.6  :      34.79  :      11.11
NEURAL NET          :           13.15  :      21.12  :       8.89
LU DECOMPOSITION    :          460.08  :      23.83  :      17.21
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 31.388
FLOATING-POINT INDEX: 31.101
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : Madison 1400MHz
L2 Cache            :
OS                  : Linux 2.6.32-5-mckinley
C compiler          :
libc                :
MEMORY INDEX        : 5.689
INTEGER INDEX       : 9.955
FLOATING-POINT INDEX: 17.250
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

Alles anzeigen

Code

processor  : 0
vendor     : GenuineIntel
arch       : IA-64
family     : 31
model      : 1
model name : Madison
revision   : 5
archrev    : 0
features   : branchlong
cpu number : 0
cpu regs   : 4
cpu MHz    : 1400.000
itc MHz    : 1400.000000
BogoMIPS   : 2097.15
siblings   : 1
physical id: 0

Alles anzeigen

Itanium 2 1500MHz 6MB L3 (Single)

Spoiler anzeigen

Code

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)


TEST                : Iterations/sec.  : Old Index   : New Index
                    :                  : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT        :          1097.9  :      28.16  :       9.25
STRING SORT         :          27.831  :      12.44  :       1.92
BITFIELD            :      2.1397e+08  :      36.70  :       7.67
FP EMULATION        :          156.47  :      75.08  :      17.33
FOURIER             :           56309  :      64.04  :      35.97
ASSIGNMENT          :          15.555  :      59.19  :      15.35
IDEA                :          1485.5  :      22.72  :       6.75
HUFFMAN             :          1344.1  :      37.27  :      11.90
NEURAL NET          :          14.087  :      22.63  :       9.52
LU DECOMPOSITION    :          490.32  :      25.40  :      18.34
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX       : 33.602
FLOATING-POINT INDEX: 33.265
Baseline (MSDOS*)   : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU                 : Madison 1500MHz
L2 Cache            :
OS                  : Linux 2.6.32-5-mckinley
C compiler          :
libc                :
MEMORY INDEX        : 6.096
INTEGER INDEX       : 10.650
FLOATING-POINT INDEX: 18.450
Baseline (LINUX)    : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

Alles anzeigen

Code

processor  : 0
vendor     : GenuineIntel
arch       : IA-64
family     : 31
model      : 1
model name : Madison
revision   : 5
archrev    : 0
features   : branchlong
cpu number : 0
cpu regs   : 4
cpu MHz    : 1500.000
itc MHz    : 1500.000000
BogoMIPS   : 2244.60
siblings   : 1
physical id: 0

Alles anzeigen

Ich hab die 1400MHz/1.5M- und die 1500MHz/6M-Konfiguration auch als Duale laufen gehabt, aber es hat sich kein wirklicher Unterschied gezeigt zur Uniprozessor-Version (dürfte ja auch net auf Dual-CPU-Konfigurationen optimiert sein).

Man merkt auch keinen Unterschied zwischen gleich hoch getakteten Prozessoren mit verschiedenen L3 Caches (1400MHz/1.5M und /4M).

mrshadowtux

Mein Raspberry Pi:

Spoiler anzeigen

pi@raspberrypi ~/pi_css5_src $ ./pi_css5 1000000
Calculation of PI using FFT and AGM, ver. LG1.1.2-MP1.5.2a.memsave
initializing...
nfft= 262144
radix= 10000
error_margin= 0.00611182
calculating 1048576 digits of PI...
AGM iteration
precision= 48: 4.20 sec
precision= 80: 4.21 sec
precision= 176: 4.19 sec
precision= 352: 4.22 sec
precision= 688: 4.20 sec
precision= 1392: 4.20 sec
precision= 2784: 4.21 sec
precision= 5584: 4.20 sec
precision= 11168: 4.20 sec
precision= 22336: 4.21 sec
precision= 44688: 4.20 sec
precision= 89408: 4.20 sec
precision= 178816: 4.20 sec
precision= 357648: 4.20 sec
precision= 715312: 4.21 sec
precision= 1430640: 4.20 sec
writing pi1048576.txt...
78.53 sec. (real time)

78,53 Sekunden für 1000000 Stellen Pi.

Mit den CFLAGS -O2 -mcpu=arm1176jzf-s -mfpu=vfp -mfloat-abi=hard

Mit -O3 statt -O2 sind es dann 78,38 Sekunden:

Spoiler anzeigen

pi@raspberrypi ~/pi_css5_src $ ./pi_css5 1000000
Calculation of PI using FFT and AGM, ver. LG1.1.2-MP1.5.2a.memsave
initializing...
nfft= 262144
radix= 10000
error_margin= 0.00611182
calculating 1048576 digits of PI...
AGM iteration
precision= 48: 4.19 sec
precision= 80: 4.19 sec
precision= 176: 4.20 sec
precision= 352: 4.19 sec
precision= 688: 4.20 sec
precision= 1392: 4.20 sec
precision= 2784: 4.19 sec
precision= 5584: 4.20 sec
precision= 11168: 4.20 sec
precision= 22336: 4.19 sec
precision= 44688: 4.20 sec
precision= 89408: 4.20 sec
precision= 178816: 4.19 sec
precision= 357648: 4.19 sec
precision= 715312: 4.20 sec
precision= 1430640: 4.20 sec
writing pi1048576.txt...
78.38 sec. (real time)

Jetzt mitmachen!