Everything in 2000

June 2000: GeForce2 MX

Shortly after GeForce 2 GTS boards were beginning to reach store shelves, NVIDIA launched another new chip, the GeForce 2 MX. Based on the GTS but designed for the corporate/mainstream market, the MX cut costs by dropping 2 pixel pipelines and reducing the clock speed putting the GeForce2 MX's fill rates at 350Mpixels/sec and 700Mtexel/sec.

GeForce2 MX

The GeForce2 MX, however, was not a simple cost-reduced version of the GTS. Two important features were added to the chip. The GeForce 2 MX incorporated TwinView, which allows a single video card to drive two separate displays, a similar feature to Matrox's DualHead technology. Second, the GeForce 2 MX added the necessary silicon revisions required to support the Macintosh platform. The PC GeForce2 MX boards reached store shelves around the same time ATI's Radeon cards did. In January 2001, the GeForce2 MX was selected as the default high-end graphics solution for the Apple Power Macintosh G4.

July 2000: Quadro2 Pro / Quadro2 MXR

Following the model set with the original GeForce 256, NVIDIA released the upmarket Quadro2 Pro and Quadro2 MXR for the performance workstation market. The Quadro2 Pro was the upmarket version of the GeForce2 GTS while the Quadro2 MXR is the upmarket version of the GeForce2 MX.

August 2000: GeForce2 Ultra

By fall, it was clear to NVIDIA that the GeForce2 GTS and GeForce2 MX would be able to hold its own ground against the ATI Radeon and the 3dfx Voodoo 4/5. This was fortunate since the NV20 wasn't going to be ready for the fall product cycle due to problems with the NV20's 0.15 micron manufacturing process and delays in the release of Microsoft's DirectX 8.

NVIDIA could afford to complete NV20 at a more leisurely pace and instead released a faster clock speed GeForce2 GTS, christened the GeForce2 Ultra.

GeForce2 Ultra

November 2000: GeForce2 Go

At the end of the year, NVIDIA launched its first mobile unit, the GeForce2 Go. Based upon the GeForce2 MX, the chip is the first mobile chip with hardware transformation, clipping, and lighting. To cut down power and cost, the GeForce2 Go only has a 720p HDTV motion compensation engine and a reduced clock speed resulting in a pixel fill rate of 286Mpixels/sec and texel fill rate of 572Mtexels/sec.

The recently announced ATI Mobility Radeon has a more advanced video accelerator with hardware iDCT which can extend battery life when watching DVDs. However, the Mobility Radeon will likely have less 3D performance than the GeForce2 Go since ATI's high-end mobile chip doesn't have T&L and only has a single pixel pipeline with three texture read units.

Although ATI has not announced any clock speeds, even with a generous estimate of a 166MHz clock speed, the Mobility Radeon will only have a fill rate of 166Mpixels/sec and 498Mtexels/sec. Of course, performance isn't everything in the mobile market where power and heat are huge issues. ATI believes that its mobile solutions will be more power efficient and the company expects strong demand for its budget mobile graphics adapters.

December 2000: NVIDIA Aquires 3dfx's Assets

After four long, hard years, the battle between NVIDIA and 3dfx finally came to an end. Unable to recover from the delay of the Voodoo5, 3dfx followed the recommendation of its Board of Directors and started the process of dissolving the company. 3dfx's first move was to sell all of its intellectual property and other assets to NVIDIA. This included 3dfx's research on anti-aliasing, Gigapixel's tile-based rendering technologies, and technology still being developed. In addition, approximately 100 former 3dfx employees joined NVIDIA bringing even more talent to the company.