Probably a rebinned 9800 that failed quality control.

Well, just after I made this post I removed the cooler, cleaned the thermal grease and took a look to the actual chip and, surprise, surprise, it's a R300, so it's totally logical that windows sees a 9700 card because it's what it is.
The mistery remains, though, on why a 9700 chip is in a 9800 board. Perhaps an OEM design, to reuse surplus, a test run of the design or something like that?

shiva2004 wrote:

Well, just after I made this post I removed the cooler, cleaned the thermal grease and took a look to the actual chip and, surprise, surprise, it's a R300, so it's totally logical that windows sees a 9700 card because it's what it is.
The mistery remains, though, on why a 9700 chip is in a 9800 board. Perhaps an OEM design, to reuse surplus, a test run of the design or something like that?

As i stated in the PM i just sent to you, mislabeled Radeons seem to be somewhat common on eBay.

I have a (Non working, missing caps) Radeon 9700 Pro that was sold to me as a X800 Pro. It even has a completely authentic looking part number label that says X800 Pro on it.

By the way it's interesting to see a different layout R300 card but I much prefer original heatsinks. Never liked alternative solution when they increase stress on weight.

I hate the noise of the originals and almost always use some better cooler, like AC or Zalman with big vent running at 5v. Even my overclocked Pentium III Xeon with XG-DLS is so silent that I hear hard drive writing instead of hurricane in my room 😀

I'm with Paadam in this, in fact I never understand why the makers put clearly insuficient coolers on the cards and then compensate making them work at very high speeds; I can begin to understand that in budget cards, but skimp on the cooling on cards that cost 300€ onward? C'mon, put a Zalman and increase the pice tag 10€ if you must.

Paadam wrote:

I hate the noise of the originals and almost always use some better cooler, like AC or Zalman with big vent running at 5v. Even my overclocked Pentium III Xeon with XG-DLS is so silent that I hear hard drive writing instead of hurricane in my room 😀

My opinion is that weight and vibrations on electric components could change dramatically switching to a new cooling system specially on a vertical card. On a cpu maybe there're already big stock dissipation system but my idea is that those cards weren't build to support such weights.

How bout removing the Zalman HSF and taking a look at the GPU?

Also look at the back of the card would be interesting.

I have quite a few Arctic passive heatpipe cooling systems with aluminium fins, they weigh not much more than standard radiator.

Regarding the weight of the Zalman, most of the weight is near the AGP connector, meaning the momentum for the weight to drag the card down is actually very small. It is easy to measure with calipers against some reference point. This only relates to the cards that have attachment points around GPU. Some coolers have mounting points elsewhere also (all over the card), that would make the momentum bigger but at the same time it makes the card more rigid.

Nowadays there is hardly any card that doesn't have 2slot huge cooler, yet they work well 😀

Paadam wrote:

I have quite a few Arctic passive heatpipe cooling systems with aluminium fins, they weigh not much more than standard radiator.

Regarding the weight of the Zalman, most of the weight is near the AGP connector, meaning the momentum for the weight to drag the card down is actually very small. It is easy to measure with calipers against some reference point. This only relates to the cards that have attachment points around GPU. Some coolers have mounting points elsewhere also (all over the card), that would make the momentum bigger but at the same time it makes the card more rigid.

Nowadays there is hardly any card that doesn't have 2slot huge cooler, yet they work well 😀

I am sure they work right. I had a complete passive heatsink replaced (fan was ko) the original Radeon 9500 Pro one but at the end its weight was absurd imho. I am not thinking there were functional problems with these solutions, only that sometimes (beside costs that probably is the main point of the stock sometimes bad solutions) original heatsink is designed in that way for many reasons, not only heat but also component stress vibrations, weight (and not only general weight, also how weight is far or close to the pcb when attached) etc.. 😉

Looking at this topic:
http://www.motherboards.org/forums/viewtopic.php?t=41681
Looks like they use the same part number for 9700 Pro and 9800 Pro.
Which is weird, but solves the mystery. It's a 9700 Pro. The best GPU of 2002. Go and have sex with it.

its not much slower than a 9800 Pro maybe 5-10%

Mistery solved, apparently Power Color sold some 9700s using the reference 9800 design, probably to reduce costs and get rid of remaining chips stock, they're unusual but nothing to write home about.
Thanks to everyone who has contributed in this thread, case closed 😎 .