GeForce FX 5600 GPU

Like GeForce FX 5800 and GeForce FX 5900, the GeForce FX 5600 family is built on TSMC’s 0.13-micron manufacturing process. In addition, both GPUs support the same 2.0+ pixel and vertex shaders NVIDIA has been boasting about since last summer. As a result, GeForce FX 5600 shares the same feature set and capabilities of NVIDIA’s flagship, albeit with lower performance.

The key difference between the 5600 series and 5800/5900 lies in its fixed function units. While GeForce FX 5900 offers four pixel pipelines with two texture units per pixel pipeline (4x2), the GeForce FX 5600 family is built on a more conservative four pixel pipeline with one texture unit per pixel pipeline architecture (4x1). This results in a reduced peak texel fill rate if the chips were clocked similarly but is largely responsible for GeForce FX 5600’s reduced transistor count: approximately 80 million in GeForce FX 5600 versus 125 million in GeForce FX 5800. With fewer transistors, GeForce FX 5600 cores are cheaper to produce for NVIDIA.

An additional difference between GeForce FX 5600 and FX 5800 lies in its memory subsystem. For GeForce FX 5600, NVIDIA has implemented twin 64-bit memory controllers; GeForce FX 5800 utilizes four 32-bit controllers while GeForce FX 5900’s memory subsystem is composed of four 64-bit controllers. GeForce FX 5600 also employs more widely available DDR memory rather than the DDR-II memory present on GeForce FX 5800.

GeForce FX 5600 models

Two GeForce FX 5600 variants are available: GeForce FX 5600 and GeForce FX 5600 Ultra. Like previous NVIDIA products, the “Ultra” product possesses higher clock frequencies than the non Ultra. Initially NVIDIA announced a core clock frequency of 350MHz for the GeForce FX 5600 Ultra, however this has recently been increased. Unfortunately, we can’t disclose the exact figures for this newer GeForce FX 5600 Ultra, but we can tell you that the GeForce FX 5600 operates at a core clock frequency of 325MHz with a memory clock of 275MHz (550MHz effective). This results in a peak fill rate of 1.3Gigatexels/sec and up to 8.8GB/sec of peak memory bandwidth.

In comparison, GeForce4 Ti 4200-8X offers a fill rate of 2.0Gigatexels/sec and 8.2GB/sec of memory bandwidth.

With its lower fill rate, GeForce FX 5600 will have a hard time competing with GeForce4 Ti 4200 in low resolution environments where texture performance is more critical. But thanks to its memory bandwidth advantage GeForce FX 5600 should be able to fare better in situations with anti-aliasing and anisotropic filtering, visual quality settings that an increasing number of gamers are taking advantage of with today’s games. GeForce FX 5600 also boasts more advanced color and Z-compression techniques – features allowing the 5600 to make more efficient use of its available memory bandwidth (again, this will play the greatest dividends in anti-aliasing and anisotropic filtering environments).

We’ve provided the following chart to summarize how the GeForce FX 5600 stacks up to the competition:

GPU	Core clock (MHz)	Memory clock	Peak fill rate (Mpixels/sec)	Peak fill rate (Mtexels/sec)	Peak memory bandwidth (GB/sec)
GeForce4 Ti 4200-8X	250	512	1000	2000	8.2
GeForce FX 5600	325	550	1300	1300	8.8
RADEON 9600 PRO	400	600	1600	1600	9.6
RADEON 9600	325	400	1300	1300	6.4
RADEON 9500 PRO	275	540	2200	2200	8.6
RADEON 9500	275	540	1100	1100	8.6