You can try wikipedia, as there is much info on that things...

"High end" is completely subjective and dependent not just on time but also target market. Just look at the Permedia2 as something that was simultaneously positioned in low-end (pro 3D market), mid range (sold a ton to Compaq) and high-end (it promised to be the best 2D/3D combo gaming card - but failed). I would replace it with something more objective like price at introduction (including where that price was)

The point is exactly to use a subjective metric. Essentially what I'm asking is "how high up the range of all available cards was this when it came out?" (Taking into account the manufacturer's positioning and where it fell wrt. the competition.) Otherwise I could just use 3Dmarks or MSRP or some other figure that's endlessly repeated elsewhere. You won't see this metric as numbers in the final visualization.

Consider
10 = bargain basement throw-in junk
100 = Bleeding-edge LAN party warrior card
200 = professional-grade unobtanium

From your example, I wouldn't call something billing itself as "the best 2D/3D combo card" to be trying for the high-end, back when it wasn't certain at all that the two features should be combined. Sounds firmly mid-range to me. I'd put it around a 70, competing with e.g. the Voodoo Rush, but I wouldn't set that in stone.

xjas wrote:

Motivation: when looking at used cards on classifieds sites I generally have no idea how they relate to each other. The model numbers all stopped making sense years ago.

I don't understand how such database could improve upon this.

And it seems you want us to do a lot of work which you could easily do yourself, rather than just fill in the gaps.

Nobody's forcing anyone to contribute if they don't want to. The google sheet is not the end product of this. I'm writing a data visualizer for my scientific work and thought it would be fun to throw it a big dataset that would be of interest to the members of this board.

There are a ton of community-driven databases of other stuff on here already, and a billion threads harping on "which card is best for this 1998 build??" I thought this would be a neat way to put some of that collective knowledge to use.

Rather than spending tens of hours researching every obsolete graphics card ever, despite others on here seemingly knowing this stuff off the top of their heads, I'm doing the work of writing the visualizer. I'm asking for help in filling in the database because I'm not an expert on it, but other board members definitely are. (This is how collaborations work in "real" science too, imagine that.) Quite frankly I don't have time or energy to take this on as a solo project; graphics cards aren't even really my thing.

Big thanks to whoever contributed the Kyro II & Voodoo4 stuff BTW. Feel free to post up & I'll give you a shoutout when the vis goes live.

See no point in doing same work again and again and again......just copy needed data from great table at Vintage3D
http://vintage3d.org/dbn.php

and let fill people data about rarity or other parameters you missing....much less work faster done....

^^ Is that database available as text or csv anywhere? I don't have the coding chops to parse a huge html table like that & extract the numbers I need.

Nevermind, I've lost all enthusiasm for this project; just no fun in it if this is the kind of reaction it generates. I'll continue developing the visualizer for my own needs & when it's done maybe I'll see about sticking Vlask's dataset into it (I didn't know that table even existed), or something like that. Or maybe not.

Thanks to those who did contribute.