Gigabyte GV-N295-18I-B rev2.0 GeForce GTX 295 Review

1st-gen GTX 295 up top, 2nd-gen bottom

Gigabyte GeForce GTX 295 and ATI Radeon 4870 X2

Armed with two GeForce GTX 200b GPUs running in tandem with one another, NVIDIA's GeForce GTX 295 is the world's fastest graphics card. But as powerful as NVIDIA's GeForce GTX 295 is, it has one key problem: the world is currently locked in a brutal year-long recession. As a result, demand for luxury items like high-end graphics cards has waned. This in turn has led to GPU price cuts, with the intent being to stir up demand.

The strategy hasn’t really worked for the most part though. GPU shipments are down and everyone’s hoping that Windows 7 and DirectX 11 will finally stoke up demand, but in the mean time price cuts continue to be the primary solution to the problem. ATI and NVIDIA just cut prices last month.

This has put the GeForce GTX 295 in a tough position. As prices for GeForce GTX 260, 275, and 285 cards continue to fall, the GTX 295’s price has remained constant. Part of the reason GeForce GTX 295 prices haven’t fallen is because of the GTX 295’s limited quantities in comparison to the other GPUs.

Another culprit though is its higher associated costs. With dual PCBs, the board’s manufacturing cost doubles, as you need twice the capacitors, voltage regulators, resistors, and other board-level components needed to outfit the board, not to mention the cost of the second PCB itself.

Both the 4870 X2 and Gigabyte GTX 295 are long cards

Note the ventilation holes on the 1st-gen GTX 295

For the past several months, NVIDIA’s been quietly concocting a solution to this problem – how do you make the GeForce GTX 295 cheaper to manufacture? Simple. You get rid of the second PCB.

Implementing a solution to this problem isn’t as simple as it sounds though. NVIDIA had to completely redesign the board to accommodate two GPUs and their associated components (rerouting traces from the GPU to memory, redesigning power delivery, etc) as well as having to develop a new heatsink/fan unit to keep everything cool. This is particularly challenging as one GPU can generate an enormous amount of heat, plopping two GPUs next to one another on a PCB you’ve now got two hotspots on the board which must be properly cooled. In fact, NVIDIA’s argued all along that dual PCB designs are superior from a cooling perspective as heat from one GPU and its power circuitry can be dissipated by its own dedicated PCB, rather than spreading heat from two GPUs across the same PCB.

On the original first-generation GeForce GTX 295 board design, a dual-slot heatsink with copper heatpipes was sandwiched in-between the two PCBs. The upper PCB was inverted so that the GPU was facing down towards the cooler. A large fan was then placed on one end of the card. This supplied fresh air to the cooler and both GPUs. The metal duct itself was also ventilated to allow hot air to escape.

It was all really rather remarkable considering the card’s 289W max board power.

NVIDIA’s second generation GeForce GTX 295 definitely looks a lot more conventional but in some ways it’s even more impressive than its predecessor. Why? Because this new GeForce GTX 295 board runs cooler and quieter.


Gigabyte sent us one of their GV-N295-18I-B rev2.0 GeForce GTX 295 cards for review. The GV-N295-18I-B rev2.0 is one of the newest second-generation GeForce GTX 295 cards to hit retail shelves. How does it compare to the original GeForce GTX 295? Let’s find out.