Is there a test that shows you the actual speed of the cpu ? Instead of just telling you how much Mhz it has ?

> Is there a test that shows you the actual speed of the cpu ?

You can use *any* computational task you like as a CPU performance benchmark. It just has to run long enough so that you can time it, or if it runs too short you'd have to run it multiple times consecutively (and for instance count how many times that task can be run in a fixed amount of time). Note that a CPU that is faster in benchmark A can be slower in benchmark B. It depends on how a benchmark's code (more precisely: the actual machine code created by the compiler from the higher level language code the benchmark is written in) fits the internal design of the CPU (or vice versa). An "actual speed of the cpu" does really only exist in terms of a specific task with fixed parameters.

Some random examples of computational benchmarks including results:

http://www.eofw.org/bench/
http://cgi.distributed.net/speed/
http://www.coremark.org/benchmark/ *
http://www.spec.org/results.html *
http://www.r-goetz.de/RGBench/bench.shtml *

(* Not available for MorphOS.)

I used this

http://aminet.net/package/util/moni/am1.001_base
http://aminet.net/package/util/moni/am1.001_mos

You can also find some benchmarks (and some "real life application" benchmarks) here. Made on Efika and several other MorphOS-capable machines (more to come soon). The site is in Polish but I guess it's pretty self-explanatory (nevertheless I might prepare an English version if needed).

Oh, If you'd like to add your results to this page - feel free to write a PM to me.

[ Edited by koszer on 2010/11/15 15:32 ]

OK, the page updated with some PowerMac G4 benchmarks made on AmiParty X in Chelm. Although I believe I need to repeat one of the tests on PMG4 2x500 MHz (I don't believe it to be slower than Efika :) ) I guess you get a quite good overview.

> I believe I need to repeat one of the tests on PMG4 2x500 MHz
> (I don't believe it to be slower than Efika :) )

Sometimes, benchmark results can be completely baffling:

"At the same frequency (400 MHz), BITFIELD on Mac G4 gives results that are a bit worse than on MPC5200. FP EMULATION is twice lower. [...] FOURRIER gives very strange results : they are better on a MPC5200 at 396 MHz than on a G5 (970FX) at 1.8 GHz !"
http://amigadev.free.fr/powerpc/nbench.html

koszer

Thank you soooo much for the link to those benchmarks. I have been looking for something like this since pmac support came out. Very intersting results! Who has that overclocked g4 board? How did they do it? tia

I would rather like to see benchmark software that runs on my machine so I can see the differences when making adjustments to ram, bus or cpu ...

And what stops you from doing all the benchmarks listed on that page (except for, perhaps IFX and Quakes)?

> http://aminet.net/package/util/moni/am1.001_base
> http://aminet.net/package/util/moni/am1.001_mos

Btw, there's version 1.002 since March 2009:

http://natmeg.stamey.at/downloads/amigamark/am_lpa2k3-morphos_update.lha

Update:

> there's version 1.002 since March 2009

New version 1.1:
http://www.219.dic.at/?pid=detail&id=ddaa92c4