Benchmarks (was Re: XScaled Squeak)

Pablo Lopez PLopez at temaiken.com.ar
Tue May 20 15:23:15 UTC 2003


(Just for reference) MY desktop P4 1.8GHz / WinXP :
			 151121605 bytecodes/sec; 3695179 sends/sec

PocketPC:
	Compaq iPAQ (PE 2030) H3630 StrongARM 266 Mhz:
			 10519395 bytecodes/sec; 390602 sends/sec
 
	HP iPAQ H5455 Intel 400 Mhz (PXA 250):
			 14925373 bytecodes/sec; 362643 sends/sec

Despite the benchmarks, my H3630 goes considerably faster than the H5455...
(cache issue?)

-----Mensaje original-----
De: Aaron J Reichow [mailto:reic0024 at d.umn.edu] 
Enviado el: Martes, 20 de Mayo de 2003 10:51 a.m.
Para: The general-purpose Squeak developers list
Asunto: Benchmarks (was Re: XScaled Squeak)


On Mon, 19 May 2003, Dan Ingalls wrote:

> As one who hasn't had a chance to use all these machines, I would find 
> it very useful if folks would include the actual tinyBenchmark results 
> for reference.  They are SO easy to cite, even if they are an 
> imperfect measure.

OK, just FYI: the Zaurus SL-5500 I had was doing 1.22M bytecodes/sec, 400K
sends/sec; the iPAQ 3100 1.0M bytecodes/sec, 350K sends/sec; the Jornada 720
1.1M bytecodes/sec, 380K sends/sec; Axim X5 at 400 MHz 1.5M bytecodes/sec,
380K sends/sec; Axim X5 at 300 MHz 1.16M bytecodes/sec, 300K sends/sec.

I am thinking of making a little image that just runs a few benchmarks, gets
MHz/Proc/RAM/bus info and adds the results to a database.  It would be
pretty straight forward, I think I'll work on it today while at work.

Other than tinyBenchmarks and macroBenchmarks, what kind of benchmarks would
people recomend being a part of this package?  It'd be nice to test morphic
speed, general display speed, event processing, but not sure if some of
these haven't been conceived as of yet.

Regards,
Aaron

  Aaron Reichow  ::  UMD ACM Pres  ::  http://www.d.umn.edu/~reic0024/ "one
has a moral responsibility to disobey unjust laws"  :: m. l. king jr.



More information about the Squeak-dev mailing list