[kwlug disc.] Benchmark discrepancies

[Khalid Baheyeldin]

> Hence, here are the scores for a Wl500gP and a WRT54GL using the
> benchmark suggested by Khalid the other day (modified for gzip rather
> than bzip2):
>

Is that because the routers do not have bzip2?

gzip is far gentler on the CPU compared to the CPU hog that is bzip2.


>
> time dd if=/dev/urandom bs=32k count=5000 | gzip -c > /dev/null &
>
> Asus WL-500g Premium (2.6.23 kernel, no wireless, idle):
>        real    19m 35.53s
>        user    0m 0.06s
>        sys     11m 8.24s
>
> WRT54GL (2.4.30 kernel, wireless on, idle):
>        real    22m 30.75s
>        user    0m 0.11s
>        sys     11m 29.45s
>

Something is a bit odd here.

If you look at the user time, it is quite close (11 sec vs 6 sec). System
time is not that far off either (11 min 8 sec vs. 11 min 29 secs).

But the sum of user and sys is off by A LOT when compared to real time (wall
clock) (11 vs. 22 mins, and 11 vs. 19 mins).

So, there may be other factors here than raw CPU/RAM performance, e.g. I/O
subsystem, random number generator?

On am AMD Athlon 64 X2, I get this:
real    1m34.546s
user    1m12.970s
sys     0m25.350s

One a Core2 Centrino Duo, I get this:
real    1m51.665s
user    1m10.012s
sys     0m42.307s

The sum of user and sys is close to real, meaning there is no wait for I/O
or anything like that.

So, I don't know what is happening there.
-- 
Khalid M. Baheyeldin
2bits.com, Inc.
http://2bits.com
Drupal optimization, development, customization and consulting.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://listserv.ccjclearline.com/pipermail/kwlug-disc/attachments/20080409/841f382b/attachment.htm