When the GF100 architecture was originally announced at Nvidia's GPU Technology Conference in November of last year, one of the promises was that the architecture was scalable. Due to its modular design, you can take away GPU clusters to scale the performance down to fill out the product line. Looks good on paper right? Of course it does! So far the GTX 480 is a card scaled down to 480 CUDA cores (down from the original 512) with the GTX 470 coming in at 448 and now the GTX 465 that comes to market with 352. To get to 352 CUDA cores you have to do more than pull a single GPU cluster that houses 128 CUDA cores in four steaming multiprocessors. You have to pull an additional streaming multiprocessor that holds 32 CUDA cores to drop you down to the magic number. With a modular design, you can do that in order to position a card into a specific performance band. Hence we now have the GTX 465.The DirectX 11 video card market has been pretty well saturated with everything from the top of the line cards, all the way down to the discrete card for OEM use. Just about every price and performance point has been covered by ATI with scant few openings left to capitalize on. One such gap was the area between the HD 5830 and HD 5850 that was left wide open, both on price and on performance. This is where the GTX465 is supposed to earn its keep. Priced at $279, the pricing point is covered so all that is left is the question of how will it perform? Well lets find out.