Specifications

The Specs

120MHz/150MHz(300Mhz) core/memory clock
Integrated geometry transform engine
Integrated dynamic lighting engine
Four-pixel rendering pipeline
480 million pixels per second fill rate
15 million triangles per second throughput
Cube-environment mapping
Single-pass emboss and dot-product bump mapping
DX6 Texture Compression
350MHz RAMDAC
2D resolution of 2048x1536 at 75Hz
AGP4X with Fast Writes
Up to 128MB SDRAM/SGRAM w/ DDR support
OpenGL ICD for Windows98, NT4, 2000
DX7 Support Powerful HDTV motion compensation.
Full frame rate DVD to 1080i resolution.

It's a GPU!

Wait, what the heck's a GPU? According to NVIDIA, a GPU (Graphics Processing Unit) is a "single-chip processor with integrated transform, lighting, triangle setup/clipping and rendering engines that is capable of processing a minimum of 10 million polygons per second." Okay, a GPU is the 3D graphics chip (with triangle setup engine and rendering pipelines) we already know and love with the addition of onboard T&L. We can't forget that a chip also has to have the ability to process 10 million triangles per second in order to meet NVIDIA's definition of a GPU. I guess we can also call the Savage2000 a GPU then…

Improvements

Our GeForce 256 DDR card has a 120MHz core and a 150MHz memory clock, but we consider the memory to be running at 300MHz because of DDR. Let's go down the list to see what's changed. Most of these changes should become standard options for all future 3D accelerators. First, let's look at the onboard T&L and quad-pixel rendering pipeline. Soon, all 3D accelerators will have to have onboard T&L and feature a rendering pipeline that can do more than two texels per clock cycle.

Cube environment mapping is a cool new feature, but we're not sure if the GeForce can actually handle the demands. It looks like texture compression is finally gaining widespread support. Compression will help keep memory bandwidth costs down while improving texture size at the same time.

The 350MHz RAMDAC isn't new, but it's an improvement from the TNT2's 300MHz RAMDAC. The V3 3500 has a 350MHz RAMDAC, and the Matrox G400 MAX has a 360MHz RAMDAC. AGP 4X with fast writes sounds like a great feature. Fast writes allow the CPU to send data directly to the GPU without having to copy the data to and from system memory. It's a shame Intel is delaying the 820 chipset for couple more months. We want AGP 4X fast writes now!