FutureMarks Horrifying Report

My impression of the GFFX line was something like this:
GFFX5800 ~= GF4Ti4600 ~= R9700 (High-end)
GFFX5600 ~= GF4Ti4200 ~= R9500 (Mid-range)
GFFX5200 ~= GF4MX ~= R9000 (Low-end)

Of course that doesn't factor in all the Ultra/Pro permutations. The comparisons are price bracket not performance btw, obviously the R9700 is faster than the Ti4600 but they are high-end cards in their generation.

 

This fits very well with the sales I see here (which I wasn't sure were representative). The TNT2 M64 used to be the common low end card, and the GeForce2 MX replaced it. The GeForce4 MX is slowly replacing it -- I guess it's not on the list because this report is too old.

 

*chuckle* I know exactly what you mean about relatives depending on you! Arggggghghghghg.... "No, Cousin, that's a mouse not a footpedal," etc.

 

Ok, I guess I am imagining my old score being there?

 

Hi guys,

Got in touch with our US office and the details about the data set that was used for the MaximumPC write up were as follows:

1) The data was derived from a sample of roughly 700k _performance analyzer_ users, not 3DMark users. To be clear about this we get two different "data feeds"; Performance Analyzer users and 3DMark users.

Performance Analyzer data is derived from systems that are analyzed when either IHV's or game companies use our online tools. I don't have specifics on which sites were used for this sample but it's most likely to consist from EA services and Nvidia based manufacturers. Another thing worth noting is that "PA users" is a more mainstream sample than 3DMark users.

2) The time frame for data collection was June 12 '02 to Sept 24 ' 02 for this sample. While I do agree that it does take quite a while for things to change among the masses and the sample is in many sense quite valid, this is nevertheless a thing worth to note.

Finally, we'll compile the first reports for 3DMark03 within following two weeks. We now have about 415.000 3DMark03 results so it should be interesting to see how things have changed. Once I have the reports, I'll open up a new thread to post the highlights here.

Cheers,

AJ

 

And the number of users accessing the "performance analyzer" is probably a more accurate depiction of the game market as a whole.

But the 3DMark01 Hall of Fame shows the most popular cards among people who actually run the benchmark:

http://www.futuremark.com/community/halloffame/

This may be a decent measure of how the DX8 and higher cards are distributed.

 

Is the performance analyzer not used as a marketing tool from IHV web sites and for other PR/MKTING activities? I think this is not really valid data as not all IHV's are using the performance analyzer equally.

Also, for the hall of fame ATI really needs to break up their product definitions in their installation inf. They take #1 with 9700/9500. Are they really top or are they #1 and #2. What about 9700 Pro, 9700, 9500 Pro, 9500???

 

Heheheh he must do that all the time now after the rage3d debacle.

 

Guys,

I sent Dave the updated data report and he was kind enough to post that in the news so it's now under another thread and can be found from here:
http://www.beyond3d.com/forum/viewtopic.php?topic=5007&forum=9

Cheers,

AJ

 

In my opinion, the pace of hardware improvement has been much faster than software's in the last year, so there are many people still playing their favourite games (I-War2 and Op. Flashpoint, for example.) on the machines they bought in 2001. I mean PIII and Gf3. This technology is still very capable, at least until DX9 games arrive (when?).

Regards
Martillo1