Amiga.org Interview : Bill Buck and Raquel Velasco (6/30/03)

[DaveP]

@Wayne

I do not EVER recall seeing BBRV saying that the G3 would be faster than the G4.

Oh he definately said it ( G3 on ArticiaS will be faster than G4 on ArticiaS ), in an attempt to justify why Genesi would not be producing G4 based cards for the Pegasos1 ( this has since changed ) after the relationship with Mai was terminated.

There were also remarks in the same paragraph that were clearly designed to "spoil" the A1XE market. Alls fair in marketing they say.

If I get idle time today, Ill go look for you.

Oh and given I now know how April works I quite understand why he made that claim.

[DaveP]

@Madgun

No need thats what "Alls fair in marketing they say." means.

@Mahen

Care to supply proof that the A1XE is "buggy" and that things need to be "deactivated"?

[DaveP]

@-D-

Well Im not sure about that the old revision Articia-S that caused April 1 to be developed for the Pegasos and the "unbranded patch" that Eyetech used on the SEs would not, from the information I have been given, have impacted overall performance of the system ( although the system would have been unreliable ).

With April 1 ( and I have no information about Eyetechs patch modus operandi or post April 1 ) the more the ArticiaS and April 1 combo get called the slower the overall system becomes because of the way April 1 is forced to patch Articia S.

So I can understand a claim that sayes "an unfixed Articia S with April 1 would actually go slower with a faster CPU" ( although that does not logically follow - it CAN go slower if the workload saturates the bus ).

However this was shortly after claiming that the new Articia S ( which claimed to fix the problem ) still had the problems - at least when used with the Pegasos.

Ergo if the new Articia S needed to use an April style fix then at least with the Pegasos the faster the maximum throughput of the CPU the more the problem gets exposed ( through the way the April handles the problem as much as the original problem itself ).

The press release extrapolates that this will be true for other chip users - like the A1-XE. We have since found out that this is not actually the case and who knows if the later April revisions do not contain a more elegant solution which solves the problem the Pegasos has with Articia-S.

Thats as good as I can go for wording without revealing sources and IP that is not mine to reveal.

Of course the caveat being if the information I have been so generously given in my search for the truth about Articia S is accurate and I have no reason to doubt it at this stage.

Our G3 Pegasos performs better than the G4 Teron with the Articia.

Which clearly, from the performance stats we have so far, this is not true for the A1XE. So unless there is something else we need to do to substantiate these claims I would suggest they are related to the pre-fixed-Articia Teron with G4.

How also this can be extrapolated to "Our G3 Pegasos running MorphOS is a faster ( per unit work ) machine than the G4 AmigaOne running Linux" I don't know.