[Zope] Re: Intel vs AMD

THoffman@indtech.wa.gov.au THoffman@indtech.wa.gov.au
Thu, 12 Dec 2002 17:38:11 +0800


Gotta add my bit.


People often give short shrift to Solaris/SPARC running python,
and pystone isn't a real world test but.

As you know python perf/pystone seems directly attributable to Mhz
so on my Dual 750Mhz Sun 280R

A single pystone is on this box is

Pystone(1.1) time for 10000 passes = 1.25
This machine benchmarks at 8000 pystones/second

Whereas 16 pystones running simultaneously on the same box
(note increased loop count to 50000 to be sure all jobs are likely to
impact each other)

Pystone(1.1) time for 50000 passes = 6.31
This machine benchmarks at 7923.93 pystones/second
Pystone(1.1) time for 50000 passes = 6.31
This machine benchmarks at 7923.93 pystones/second
Pystone(1.1) time for 50000 passes = 6.33
This machine benchmarks at 7898.89 pystones/second
Pystone(1.1) time for 50000 passes = 6.31
This machine benchmarks at 7923.93 pystones/second
Pystone(1.1) time for 50000 passes = 6.31
This machine benchmarks at 7923.93 pystones/second
Pystone(1.1) time for 50000 passes = 6.31
This machine benchmarks at 7923.93 pystones/second
Pystone(1.1) time for 50000 passes = 6.3
This machine benchmarks at 7936.51 pystones/second
Pystone(1.1) time for 50000 passes = 6.31
This machine benchmarks at 7923.93 pystones/second
Pystone(1.1) time for 50000 passes = 6.31
This machine benchmarks at 7923.93 pystones/second
Pystone(1.1) time for 50000 passes = 6.32
This machine benchmarks at 7911.39 pystones/second
Pystone(1.1) time for 50000 passes = 6.33
This machine benchmarks at 7898.89 pystones/second
Pystone(1.1) time for 50000 passes = 6.33
This machine benchmarks at 7898.89 pystones/second
Pystone(1.1) time for 50000 passes = 6.33
This machine benchmarks at 7898.89 pystones/second
Pystone(1.1) time for 50000 passes = 6.33
This machine benchmarks at 7898.89 pystones/second
Pystone(1.1) time for 50000 passes = 6.33
This machine benchmarks at 7898.89 pystones/second
Pystone(1.1) time for 50000 passes = 6.35
This machine benchmarks at 7874.02 pystones/second

So what we see is a single instance python is not really running as
fast as it could possibly, and that the increase in load for each new
process is not linear. If the tasks where running as fast as the CPU can
do the job, then 4 pystone jobs should each take twice as long as two.

So whilst solaris may not be the fastest pystone box for MHz/$ you may
find you can do far more with it if you run a bunch of Zeo servers ;-)

If you are running lot's of Zope servers then you need to look at what
your horizontal scalability will be. (ie how does it handle a load from
multiple processes.) This test obviously doesn't stress memory.

See ya

Tim


On Wed, 2002-12-11 at 22:55, Roel Van den Bergh wrote:
> Here are some stats to add :-)
> 
> Dell PowerApp 120 Intel Pentium III (Coppermine) Dual 1Ghz, 1GB PC133 ECC
> SDRAM
> running in single proccessor mode
> 
> [root@root]# python /home/zope/2-5-0/lib/python2.1/test/pystone.py
> Pystone(1.1) time for 10000 passes = 0.87
> This machine benchmarks at 11494.3 pystones/second
> [root@root]# python /home/Python-2.1.3/Lib/test/pystone.py
> Pystone(1.1) time for 10000 passes = 0.86
> This machine benchmarks at 11627.9 pystones/second
> [root@root]# python /usr/lib/python1.5/test/pystone.py
> Pystone(1.1) time for 10000 passes = 0.85
> This machine benchmarks at 11764.7 pystones/second
> [root@root]# python /usr/lib/python2.1/test/pystone.py
> Pystone(1.1) time for 10000 passes = 0.86
> This machine benchmarks at 11627.9 pystones/second
> [root@root]# python /usr/local/lib/python2.1/test/pystone.py
> Pystone(1.1) time for 10000 passes = 0.87
> This machine benchmarks at 11494.3 pystones/second
> [root@root]# python /usr/src/redhat/BUILD/Python-2.1.3/Lib/test/pystone.py
> Pystone(1.1) time for 10000 passes = 0.85
> This machine benchmarks at 11764.7 pystones/second
> 
> Roel.
> 
> 
> 
> _______________________________________________
> Zope maillist  -  Zope@zope.org
> http://lists.zope.org/mailman/listinfo/zope
> **   No cross posts or HTML encoding!  **
> (Related lists - 
>  http://lists.zope.org/mailman/listinfo/zope-announce
>  http://lists.zope.org/mailman/listinfo/zope-dev )



DISCLAIMER: This email, including any attachments, is intended only for use
by the addressee(s) and may contain confidential and/or personal information
and may also be the subject of legal privilege. Any personal information
contained in this email is not to be used or disclosed for any purpose other
than the purpose for which you have received it. If you are not the intended
recipient, you must not disclose or use the information contained in it. In
this case, please let me know by return email, delete the message
permanently from your system and destroy any copies. Emails and their
attachments may be interfered with, may contain computer viruses or other
defects and may not be successfully replicated on other systems. All
attachments are opened at the recipient's risk.