GeForce_GTX_280_3qtr

GeForce_GTX_280_3way 

Think general purpose microprocessors have high transistor counts? Think again folks! NVIDIA has unveiled their latest family of GeForce GTX 200 GPUs and these bad boys have a TON of transistors - 1.4 billion transistors to be roughly exact. Making up the GTX 200 family are two processors - the high end GTX 280 and the mid-range GTX 260. The new GPUs are basically the same processor with the GTX 280 having all the capabilities enabled and the GTX 260 having some of the capabilities disabled.

With the new GTX 200 series of GPUs, NVIDIA is stating that bigger is indeed better - at least for the high end graphics world. They’ve packed a ton of horse power into a single die - resulting in a processor which is bound to be super large and power hungry. Thus - don’t expect graphic cards using the GTX 200 processors to be small in size. Heck you might even need a PSU upgrade.

Now contrast this with ATI’s philosophy which boils down to multi-processors for scalability. They’re keeping their designs small and fairly compact - hoping to tack on more processors if the demands are needed. In fact, AMD has pretty much conceded the high end graphics market to NVIDIA, instead focusing on delivering near high end performance at mid-range costs.

With the new GTX 280 based cards expected to cost around $650 and GTX 260 cards around $400, one has to wonder if these processors are worth it? Well, the reviews thus far have been somewhat mixed. One thing is certain - the GTX 200 is definitely the fastest processor even designed by NVIDIA. However, many of also found that current 9800 GX2 cards are just as fast if not faster than the single GPU GTX 280. Another interesting observation found in the reviews: The GTX 260 is the better bargain between the two new GPUs simply because you get 15-25 % performance degradation at a 40% cost reduction.

For more “light” reading, check out the early reviews from:

Technorati Tags: , ,