[Advcomparch] Benchmarks

Stratos Psomadakis psomas at ece.ntua.gr
Thu Apr 16 01:23:44 EEST 2009


τα benchmarks είναι ένα πρόγραμμα που κάνει compress το input
του(bzip2), ένα πρόγραμμα που λύνει το πρόγβλημα της μέγιστης ροής σε
ένα δίκτυο νομίζω(mcf), και ένα πρόγραμμα για σκάκι(crafty)...

αυτό που λες συμβαίνει γτ από το script(το runbenchs.simics) λείπει μια
γραμμή:
instruction-fetch-mode instruction-fetch-trace

by default(μάλλον έχει να κάνει και με τα internal caches που
χρησιμοποιεί το simics, αν και δεν είμαι σίγουρος) το
instruction-fetch-mode είναι τέτοιο ώστε τα
instruction fetches να μην είναι ορατά στο memory hierarchy(στα cache
modules που προσθέτεις)...
με αυτό σε μένα έπαιξε σωστά(τα stats και τα περιεχόμενα της ic
φαίνονταν normal)...
O/H Thomas Pazios έγραψε:
> Ta benchmarks ti diafores exoun metaksy tous?
>
> Etreksa to bzip2 kai to mcf kai mou evgalan poly paromoia apotelesmata, enw h
> IC exei ola ta statistika ths mhdenika kai sta 2 benchmarks
>
>
> Parathetw ta apotelesmata tou bzip2 gia mnhmh 32K kai associativity 4 kai line
> size 64
>
>
>
>
> Statistics for cpu cpu0
>        User  Supervisor       Total  Description
>     4520860 14217049586 14221570446  instructions executed
>     1767088    52822054    54589142  memory read operations
>     1064679    31071111    32135790  memory write operations
>           0     4804035     4804035  I/O read operations
>           0        8882        8882  I/O write operations
> processor        steps       cycles  time [s]
> cpu0       14221570446  14510913369   725.546
> Simulating...
> Collect statistics
>
> Cache statistics: dc
> -----------------
>     Total number of transactions:     432233153
>
>          Device data reads (DMA):             0
>         Device data writes (DMA):             0
>
>           Uncacheable data reads:           933
>          Uncacheable data writes:           816
>  Uncacheable instruction fetches:             0
>
>           Data read transactions:     272106116
>                 Data read misses:         73274
>              Data read hit ratio:         99.97%
>
>   Instruction fetch transactions:             0
>         Instruction fetch misses:             0
>
>          Data write transactions:     160125288
>                Data write misses:         20529
>             Data write hit ratio:         99.99%
>
>           Copy back transactions:             0
>
>                Lost Stall Cycles:       5858731
>
> Cache statistics: ic
> -----------------
>     Total number of transactions:             0
>
>          Device data reads (DMA):             0
>         Device data writes (DMA):             0
>
>           Uncacheable data reads:             0
>          Uncacheable data writes:             0
>  Uncacheable instruction fetches:             0
>
>           Data read transactions:             0
>                 Data read misses:             0
>
>   Instruction fetch transactions:             0
>         Instruction fetch misses:             0
>
>          Data write transactions:             0
>                Data write misses:             0
>
>           Copy back transactions:             0
>
>
> Cache statistics: l2c
> -----------------
>     Total number of transactions:     160200311
>
>          Device data reads (DMA):             0
>         Device data writes (DMA):             0
>
>           Uncacheable data reads:           933
>          Uncacheable data writes:           816
>  Uncacheable instruction fetches:             0
>
>           Data read transactions:         73274
>                 Data read misses:           651
>              Data read hit ratio:         99.11%
>
>   Instruction fetch transactions:             0
>         Instruction fetch misses:             0
>
>          Data write transactions:     160125288
>                Data write misses:           257
>             Data write hit ratio:        100.00%
>
>           Copy back transactions:             0
>
>                Lost Stall Cycles:       2928830
>
> Statistics for cpu cpu0
>        User  Supervisor       Total  Description
>     4605845 15216964601 15221570446  instructions executed
>     1790403   324922247   326712650  memory read operations
>     1077432   191183646   192261078  memory write operations
>           0    29064671    29064671  I/O read operations
>           0       58686       58686  I/O write operations
> processor        steps       cycles  time [s]
> cpu0       15221570446  18401337040   920.067
> [cpu0] cs:0xc01fa3ef p:0x001fa3ef  pop ebx
>
>
>
>
>
> ta apotelesmata tou mcf einai paraplhsia se megalo vathmo.
>
> _______________________________________________
> Advcomparch mailing list
> Advcomparch at lists.cslab.ece.ntua.gr
> http://lists.cslab.ece.ntua.gr/mailman/listinfo/advcomparch
>   


-- 
Stratos Psomadakis
<psomas at ece.ntua.gr>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 260 bytes
Desc: not available
URL: <http://lists.cslab.ece.ntua.gr/pipermail/advcomparch/attachments/20090416/601f24ef/attachment.pgp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 260 bytes
Desc: OpenPGP digital signature
URL: <http://lists.cslab.ece.ntua.gr/pipermail/advcomparch/attachments/20090416/601f24ef/attachment-0001.pgp>


More information about the Advcomparch mailing list