But what’s new about the GeForce FX? Well, frankly, not a whole lot. Despite the massive transistor count, no less than 125-million, the GeForce FX brings only a few new features to the table, the first being the much touted CineFX Engine. The CineFX engine is just another pretty name for the GeForce FX’ vertex and pixel shaders and the addition of DirectX 9.0 support, no more, no less. Intellisample Technology is the next buzzword which basically covers Nvidia’s new anti-aliasing and anisotropic filter implementation in the GeForce FX, and there’s frankly also nothing revolutionary about it, just another evolutionary step over the previous generation. The one feature that does seem to weigh in heavily on the ‘new and innovative features’ scale is the use of DDR-II memory running at no less than 500MHz clockspeed which yields an effective clockspeed of 1GHz. Nvidia chose to use a 128-bit wide databus for this new memory architecture, whereas ATi's Radeon 9700 Pro uses 256-bit. We'll have to see though how effective their DDR-II implementation is and whether it has enough raw bandwidth to keep the GeForce FX' core running at full speed.
Fig 2. A closer look at the DDR-II memory chips from Samsung and the Molex power connector.
Another first is the 0.13-micron core of the GeForce FX, as although they packed 125-million transistors onto the GeForce FX it is relatively compact. Unfortunately it does consume a whole lot of power and needs an external Molex power connector as the power requirements are beyond the AGP spec. With a power consumption of up to 50-watts under full load the GeForce FX puts many desktop CPUs to shame and needs a cooling solution that’s more reminiscent of a hair dryer than anything else. Unfortunately the same can be said about the noise production of this rather elaborate cooling solution, as when the 3D pipeline starts to crank out those pixels at full steam the heatsink’s fan also speeds up to higher RPMs and starts to sound much alike a mini hair dryer. If that gets you excited and brings up fond memories of CPUs overclocked to twice their original clockspeed your excitement will soon be tempered by the fact that the GeForce FX’ GPU 'only' runs at a 500MHz clockspeed, not a whole lot more than previous generations that featured no such elaborate cooling solutions and not what you’ve come to expect from a die-shrink down to 0.13-micron.
Fig 3. The FX Flow cooling solution of the GeForce FX, taking up the AGP and the adjacent PCI slot.
The other disadvantage is that the cooling solution takes up two slots, the AGP and the adjacent PCI slot, not because we’ll get triple DVI or video-in/out, but because the cooling solution needs the 2nd slot for air intake, how’s that for a change? Nvidia pitches it as a ‘radical and revolutionary patent pending dynamic thermal management solution, that’s an enthusiasts dream and gives gamers the right to brag’ we’ll just call it like it is; a bulky, noisy heatsink that could probably double as a hair dryer and really does nothing to impress us. If there’s one thing to brag about it is the fact that the noise of the fan will probably drive people nuts faster than any other videocard before this one, whether that’s something positive we’ll leave up to you.
Discuss This Article (12 Comments) - If you have any questions, comments or suggestions about the article and/or its contents please leave your comments here and we'll do our best to address any concerns.
Rate This Product - If you have first hand experience with this product and would like to share your experience with others please leave your comments here.