Indianapolis (IN) - When this card was purchased new, I paid $565 for it. At some point, the card was damaged and no longer functions. BIOS won't recognize it. The fan doesn't spin. Nothing. EVGA tells me it's outside their warranty timeframe, so I thought it might be fun to see what makes it tick. See the slideshow and find out what goes on inside those slick looking video card packages. You might be surprised how small the actual video card is in there.





Final dimensions



When fully disassembled, the video card is only 7 mm thick, excluding things which jut out like capacitors and physical connectors. In fact, were it not for the G80-300 GPU itself, the board would only be about 4 mm thick. It weighs just over one pound when everything is removed and has a real heft to it when held - much more so than I was expecting.



For anyone who's worked with modern Intel or AMD CPUs, the physical package of the NVIDIA G80-300 GPU is close to the same size, maybe slightly larger. The GDDR chips were surprisingly small. This card has 768 MB of GDDR3 memory, and yet there are only 12 physical chips, each of which is about half the size of a postage stamp. For some reason, I would've expected more chips on the board.

Tricky disassembly


Disassembling the card was a little tricky, though not a huge deal. It did take some effort to separate the heatsink/fan from the card itself. I may have actually damaged the card (were it not already damaged) with the twisting force required to separate the heatsink/fan. I think with a little more patience I could've gotten it apart just by wiggling. As it was already broken, I didn't mind that prying was on the menu.



All in all there are four main parts. The outer shroud, the heatsink/fan assembly, the card itself, and a strange aluminum (what looks like a harness) surrounding the GPU. The harness is slightly larger than the GPU and has eight screws which affix it from the back side.

The harness is shorter than the GPU package itself, and doesn't appear to touch anything when assembled. I can only assume it's there to provide structural strength, though I'm not sure why it would need to be there once the heatsink/fan is seated.

Its size, unusual shape and location can be seen etched into the PCB around the GPU on slide 5 and along the edges on slide 7. Also, if you look carefully on slide 6 in the middle where the X is, you can see the multiple screw hole connectors which affix it from the back.

---

---

The slideshow includes some close up pics of the on-board chips, as well as indicators where things are bolted, where contact is made between the board and heatsink/fan, and more. Enjoy!