[citation][nom]Crashman[/nom]You speak of two single-slot cards getting the inside of the case hot and have a good point, since those don't exhaust their own heat. However, your commend about "melting the thing" with a card that does exhaust its own heat is nonsense. Seriously, it doesn't make any sense whatsoever. For example, you'd be better off with an HD 4850 X2 (Sapphire makes them) BECAUSE it exhausts its own heat than you would with two HD 4850 single-slot cards.Furthermore, the system WAS retested with a GTX295, read the conclusion.[/citation]
First off, you know as well as I do that the "melting the thing" is an exaggeration.
I did miss that sentence in the conclusion (stating power consumption and temps) - I recognize the data collected was mostly on the 260 aside from that brief interlude and the bit at the beginning about size.
However, you can't tell me that the GTX 295 is not going to be recirculating air into the case. I can't speak for the Radeon, as I have not used the 4870x2, but unless you tape up the massive side vent on the top of the card, it's going to be recirculating air. Now, in
my experience with a well ventilated case, that's enough to cause heat problems with the rest of the components. Perhaps the GTX 295 had a redesign on their reference cooler that I wasn't aware about, or the case was open when this was tested (for some silly reason), so that the recirculated air did not matter as much? Or maybe my experience has been tainted by a bad card? Could be a number of things. However, I'm not going to deny my personal experience:
1: The dual slot reference cooler for the dual pci cards (nvidia) recirculates air into the case, heating up all internal components
2: SFF cases, by their nature (whether it be constricted air space or lack of cable management), have an overall higher internal temp.
I'm betting that the temps they took to get the 44degC idle? It's an open air test, no case - or case with the side panels removed, etc.