home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Amiga Plus 1995 #1
/
Amiga Plus 1995 #1.iso
/
testprogramme
/
aibb_6.5
/
documentation
/
!readme.important
< prev
next >
Wrap
Text File
|
1994-12-13
|
3KB
|
51 lines
A.I.B.B.
Amiga Intuition Based Benchmarks
Program Release Version 6.5
Copyright 1991-1993 LaMonte Koop
All Rights Reserved
Recent updates to both the Amiga's hardware and operating system make
careful attention necessary when doing performance comparisons. The
most prevalent items of attention which can affect a systems rating are the
new AGA graphics chips, and the inclusion of AmigaOS 3.0 (and eventually
above) to the Amiga.
The AGA chipset is now being shipped in both the A4000 and A1200, and
will most likely be included in some form or another in future Amiga
models. This chipset is capable of increased graphics bandwidth, and thus
will run graphics-intensive tests at faster rates than a similarly
processor-equipped ECS or older chipset based Amiga. Moreover, the
AGA set has mandated architectural changes to CHIP RAM, and how the
custom chips act upon it. Because of this, it is important to take these
factors into consideration when noting benchmark results between AGA and
non-AGA equipped systems.
In order to take advantage of the AGA chipset, AmigaOS 3.0 is being
utilized on systems utilizing such. This version of AmigaOS is not
restricted to use on AGA systems only, however. For this reason, it is
very important that on tests which utilize OS functions (most prevalently
the graphics tests within AIBB) that said OS version be taken into account.
AmigaOS 3.0 is faster and more finely optimized in many ways than previous
OS releases. Graphics routines within often run much faster than similar
counterparts on pre-3.0 releases, and will thus reflect a higher rating
in test measurments. This MUST be considered when making any comparison
between pre-3.0 based machines and ones utilizing AmigaOS 3.0 and above.
The above proves to illustrate that MANY factors must be considered to
make proper comparisons between any two systems, let alone many. If a test
result seems different than what would be expected, check ALL the parameters
which may differ. AIBB itself contains both systems running 3.0 and
2.04 OS versions in its default comparison machines, and this should not be
overlooked when making a good comparison with a given host system. At
present, the only 3.0 equipped system in the default lineup is the A4000,
but the influence of using AmigaOS 3.0 on system performance (mostly in
the areas of tests which use graphics calls, and other system kernel calls
-- stock tests which do not directly use the OS are not affected) should
not be ignored.
In addition to all this, AIBB also now allows graphics tests to be
performed in a user-selectable screen mode on systems running AmigaOS 2.1
or greater. This fact necessitates that comparisons be checked VERY
carefully, as differing screen modes can influence performance results.
AIBB will attempt to give warning if compared systems were using differing
screen modes, but ultimately it is up to the person making the judgements
to be cautions about this.