Need AES benchmark on Intel 64 bit

Need AES benchmark on Intel 64 bit

[Andreas Steinmetz]

[-- Attachment #1: Type: text/plain, Size: 727 bytes --]

Hi,
can anybody help out? I don't have access to Intel 64 bit CPUs and need
some microbenchmark results on Intel 64 bit. Usage guide for the
attached archive:

'ref' contains the current generic AES implementation
'new' contains the 64 bit AES assembler implementation

Do 'make' in both directories and run the resulting 'aes' on an
otherwise idle system without any cpufreq (speedstep) stuff active.
Preferrably do multiple runs to assert that the results are usable (a
few ticks difference between runs are ok).

The microbenchmark is set up to produce somewhat real life results with
hot caches, thus the same data block is processed all the time.
-- 
Andreas Steinmetz                       SPAMmers use robotrap@domdv.de

[-- Attachment #2: microbench.tgz --]
[-- Type: application/x-tar-gz, Size: 6230 bytes --]

Re: Need AES benchmark on Intel 64 bit

[Wes Felter]

Andreas Steinmetz wrote:
> Hi,
> can anybody help out? I don't have access to Intel 64 bit CPUs and need
> some microbenchmark results on Intel 64 bit. Usage guide for the
> attached archive:
> 
> 'ref' contains the current generic AES implementation
> 'new' contains the 64 bit AES assembler implementation

This is on a 3.6GHz Nocona:

wmf:~/src/aes/ref> ./aes
schedule128 1213
schedule192 1443
schedule256 1736
enc asm 128 1020
dec asm 128 1049
enc asm 192 1352
dec asm 192 1387
enc asm 256 1541
dec asm 256 1578
wmf:~/src/aes/ref> ./aes
schedule128 1214
schedule192 1442
schedule256 1736
enc asm 128 1021
dec asm 128 1049
enc asm 192 1352
dec asm 192 1388
enc asm 256 1541
dec asm 256 1579

wmf:~/src/aes/new> ./aes
schedule128 1276
schedule192 1520
schedule256 1822
enc asm 128 790
dec asm 128 795
enc asm 192 936
dec asm 192 937
enc asm 256 1087
dec asm 256 1086
wmf:~/src/aes/new> ./aes
schedule128 1276
schedule192 1520
schedule256 1821
enc asm 128 791
dec asm 128 796
enc asm 192 936
dec asm 192 937
enc asm 256 1087
dec asm 256 1086

Wes Felter - wesley@felter.org

Re: Need AES benchmark on Intel 64 bit

[Dave Jones]

On Thu, Apr 21, 2005 at 02:28:54PM +0200, Andreas Steinmetz wrote:
 > Hi,
 > can anybody help out? I don't have access to Intel 64 bit CPUs and need
 > some microbenchmark results on Intel 64 bit. Usage guide for the
 > attached archive:
 > 
 > 'ref' contains the current generic AES implementation
 > 'new' contains the 64 bit AES assembler implementation
 > 
 > Do 'make' in both directories and run the resulting 'aes' on an
 > otherwise idle system without any cpufreq (speedstep) stuff active.
 > Preferrably do multiple runs to assert that the results are usable (a
 > few ticks difference between runs are ok).
 > 
 > The microbenchmark is set up to produce somewhat real life results with
 > hot caches, thus the same data block is processed all the time.

ref:
schedule128 1535
schedule192 1817
schedule256 2167
enc asm 128 1166
dec asm 128 1163
enc asm 192 1361
dec asm 192 1378
enc asm 256 1570
dec asm 256 1594

schedule128 1535
schedule192 1817
schedule256 2170
enc asm 128 1166
dec asm 128 1164
enc asm 192 1361
dec asm 192 1382
enc asm 256 1579
dec asm 256 1599


new:
schedule128 1542
schedule192 1823
schedule256 2177
enc asm 128 1034
dec asm 128 1018
enc asm 192 1212
dec asm 192 1213
enc asm 256 1409
dec asm 256 1401

schedule128 1547
schedule192 1828
schedule256 2182
enc asm 128 1036
dec asm 128 1024
enc asm 192 1217
dec asm 192 1222
enc asm 256 1412
dec asm 256 1395


subsequent runs present the same picture. 'new' seems slightly higher
scores on schedule*, but lower scores on the other tests.


Dual 2.8 Xeon w/HT...

processor       : 3
vendor_id       : GenuineIntel
cpu family      : 15
model           : 4
model name      :                   Intel(R) Xeon(TM) CPU 2.80GHz
stepping        : 1
cpu MHz         : 2793.079
cache size      : 1024 KB
physical id     : 3
siblings        : 2
fpu             : yes
fpu_exception   : yes
cpuid level     : 3
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm syscall nx lm pni monitor ds_cpl cid cx16 xtpr
bogomips        : 5570.56
clflush size    : 64
cache_alignment : 128
address sizes   : 36 bits physical, 48 bits virtual
power management:


		Dave