>Err.. I wish I still had the article, but the Pentium 90Mhz tested lower than
>the 601/60... The only thing it was faster on was integer math, and there it
>was slower than the pentium 60.
From the research that I have done for my Mac & IBM compare sheet I know this
comment to be totally bogus.
Ingram 94 report
Machines tested: Power Macintosh 6100/60, 7100/66, and 8100/80
Compaq Deskpro Pentium/60 and Pentium/66, and Compaq Presario 486SX/25
All machines had 16 MB of RAM. Standard configurations
Programs: 25 tests incorporating Adobe Photoshop, Aldus Freehand, Frame
Technology's FrameMaker, Fractal Design Painter, and some other programs
available on both platforms
Tasks: included opening files, scrolling, and spell checking
Conclusions:
1) Power Macintoshes were $1,000 cheaper than comparitive Pentium boxes.
{at the time of the study}
And that is ignoring the built-in SCSI-2, networking, and audio features
that gererally cost extra for most Intel based machines.
2) Power Macintosh 6100/60 performed 24% faster than a Compaq Deskpro
Pentium/60
3) Power Macintosh 6100/60 performed 5% faster than a Pentium/66
4) Power Macintosh 7100/66 performed 38% faster than a Pentium/66
5) Power Macintosh 8100/80 performed 54% faster than a Pentium/66
A rough calculation indicates that a Power Macintosh 8100/80 would be ~12%
faster than a Pentium 90 if everything was equal.
A Computer Shopper, June 1994 v14 n6 p146(12) article showed that a 8100/80
and Pentium/90 actually ran at about the same speed.
Given this information there is no way a Pentium 90Mhz could "test lower than
the 601/60". The math is simple:
Given
Power Macintosh 6100/60 = 1.05 x Pentium/66
Power Macintosh 8100/80 = 1.54 x Pentium/66
Power Macintosh 8100/80 ~= Pentium/90
Therefore Power Macintosh 8100/80 =~ 1.47 x Power Macintosh 6100/60
and Pentium/90 =~ 1.47 x Power Macintosh 6100/60
Better optimized code for the Pentium explains a greater than MHz {1.37 times}
increase of the Pentium/90 over the Pentium/66.
[Chop!]
> Tasks: included opening files, scrolling, and spell checking
> Conclusions:
[Chop again]
> Given this information there is no way a Pentium 90Mhz could "test lower than
> the 601/60". The math is simple:
> Given
> Power Macintosh 6100/60 = 1.05 x Pentium/66
> Power Macintosh 8100/80 = 1.54 x Pentium/66
> Power Macintosh 8100/80 ~= Pentium/90
Oops... those tasks you list are not exactly what you would call floating
point intensive. Although I agree that the PM6100/60 is not faster in
everyday tasks than the P5-90 it MAY be faster on Floating point intensive
operations.
It is hard to really take advantage of the PowerMac speed right now since
much of the software is still emulated. This means that until much more
native software is available us poor PowerMac users are not even close to
seeing the potential of our machines realized (MS Excel is a real drag).
But when you consider that the BEGINNING of the PowerPC line is competing
in speed with the "Pentium" of the Intel line, it spells a VERY sorry
future for Intel. And the funny bit is they know it. After all you have
to remember that the PowerMac 6100/60 that they are so casually comparing
to Pentium boxes will be the SLOWEST PowerMacintosh ever made :-)
<snip>
>>the 601/60... The only thing it was faster on was integer math, and there it
>>was slower than the pentium 60.
>From the research that I have done for my Mac & IBM compare sheet I know this
>comment to be totally bogus.
I wouldn't say totally bogus, i watched a 66mhz Powermac 16meg ram
perform a series of FPU intensive filters in Photoshop (native) 3
TIMES faster than a Pentium 60 (compaq) running a specially
compiled version of Photoshop. The 486 DX2/66 was only half way through
the test.
In the integer performance i would agree the Pmac is as fast or a little
bit faster than pentium. In fpu performance the PPC is waaaaaaaay ahead.
I prefer real world tests myself
Ok, time to killfile bgrubb again. sigh, i thought i had left this guy
behind
byeeee
Jeremy Mears
je...@cairo.anu.edu.au