Temp (Armand.Hirt@caramail.com) on 12/4/05 wrote:
---------------------------
>
>No, and while you have repeated this a couple of times now, I think your past attempts
>at rationalizing it have been rebutted sytematically by references to the post.
>I asked if "the availability of the system [was] explicitely tied to the release
>of [the published] benchmark figures", which does not imply it actually was. I immediately
>followed with: "Or were you free to choose any benchmark?", which of course does not imply David's hands were tied.

The entire line of questioning started with:

"What motivated you in this choice of applications? From reading the article, I get the impression that the system came together with Intel-recommended benchmarks, that you installed and ran. Is that what happened?"

Which was then followed by the additional questions. I can pick any specific quote out of any specific post and 'prove' what I want out of context. In this case, the context was "Did you just run what Intel recommended or required?"

>
>How can one claim that the former implies a "yes", while the latter implies a "no"?

'plausible deniability'

>It seems to me that one could as well claim that I "suggested/implied" that David
>had been free to choose any benchmark. Of course, it would be equally invalid, since
>there was no suggestion in one or the other direction.

Heh. Whatever.

>
>Framkly, that makes one wonder if the app does check for CPU brand ID in order
>to get to know whether it supportts SSE :-)

If I had an application that was fairly popular, and Intel came to me and said "I would like to pay you to optimize your application for my processor, and I will provide the expertise as well", I would be foolish to say no. Intel owns 80% of the market, so getting paid to optimize for it very likely makes good business sense, even if it doesn't help or even slightly 'penalizes' AMD based systems. For Intel, it also makes good business sense. OTOH, if AMD did the same thing it may or may not make good business sense depending upon whether it would be detrimental in any way to those running the app on Intel based systems.

If that app is fairly popular, and I am trying to evaluate what a typical user will see, then it makes sense to use that application whether suggested by Intel or not. If I use only that app, then it is a crappy review. If I use several other apps, some of which have not been optimized, then users can see the range of performance they can expect. If I can further categorize the apps by what their resource usage/bottlenecks are, and how they are optimized, I can then make it an even more useful review by offering various application 'profiles' for users to try and estimate the benefit with their own apps.

If I use only applications that are not optimized for Intel, whether or not they are used in the field, that would be a crappy review as well, since there are obviously at least several apps optimized in this way (and likely not so much for AMD). This, of course, would make sense from a market perspective, since software houses are more likely to optimize for the dominant implementation.

If the app happens to check for CPUID, or some set of bits - that might be nice to know in a comparision between two manufacturer's products - but in the end, if that is how the app runs in the field then that is what the user will actually experience, true? At that point, it would likely be a good idea to test with some apps that *don't* check CPUID, and provide the results of those as well. This would be particularly pertinent in a head-to-head shootout review... but not so much in the one being discussed, don't you think?

Once again, we go back to my original position - it is better to understand the reason why an app was selected than to just spread innuendo and doubt. The former indicates an open minded investigation, the latter suggests an agenda.

Regards,
Dean

>
>- Armand