NVIDIA Corp. broadened its already expansive graphics line with the introduction of four new NVIDIA Quadro professional graphics solutions based on PCI Express. Leveraging this next-generation bus architecture, NVIDIA doubles the bandwidth of its AGP 8X-based products to more than 4GB per second in both upstream and downstream data transfers.

With the new NVIDIA Quadro FX 3400 high-end graphics model, NVIDIA enables system builders to assemble the ultimate scalable, 64-bit powerhouse by configuring multiple graphics subsystems in a single workstation. NVIDIA's new professional graphics hardware will be widely available in workstations from Dell Computer, Fujitsu-Siemens Computers and HP, among others.

"We've been accustomed to multiprocessing systems for sometime with CPUs, but it is only with the industry's transition to PCI Express that we're seeing multiple graphics processing units in a single, affordable, mainstream workstation," said Anthony Kros, associate analyst of workstations at Gartner Inc. "Multiprocessing graphics scalability-combined with a 64-bit platform and the extra bandwidth PCI Express inherently brings to the table-represents a colossal surge forward for high-performance computing."

"Over the years, we've gone from offering a couple of AGP-based 3D graphics solutions, to offering a wider range of 3D AGP graphics solutions, a 3D PCI solution, 2D professional graphics solutions, digital media solutions and even multiheaded options," added Jeff Brown, general manager of professional graphics at NVIDIA. "With today's introduction of 2D and 3D PCI Express-based graphics, we can say with absolute certainty that we have something powerful and appropriate for every kind of professional user."

The new breadth of solutions based on PCI Express includes the NVIDIA Quadro FX 3400, FX 1300 and FX 330 for 3D applications, and the NVIDIA Quadro NVS 280 for professional 2D applications. In addition to multiprocessing scalability and increased bandwidth, these solutions feature:
* Support for NVIDIA Scalable Link Interface (SLI) hardware and software, allowing users to plug two NVIDIA Quadro graphics boards into their PCI Express-compatible motherboards;
* Increased application performance;
* Advanced vertex and pixel programmability to allow for the creation of realtime shaders that can simulate a range of physical effects and surface properties;
* True 128-bit floating-point precision, which enables mathematical computations that maintain higher accuracy, resulting in greatly enhanced visual quality;
* High-precision dynamic-range technology for extended 32-bit floating-point capabilities in shading, filtering, texturing and blending operations;
* Blazing fast pixel read-back performance for faster graphics-to-system throughput; and
* Complete compliance with NVIDIA Unified Driver Architecture, a set of backward-and-forward, top-to-bottom compatible graphics drivers that optimize application performance, stability and reliability.

"The enormous bidirectional bandwidth of the PCI Express-based NVIDIA Quadro FX family allows us, for the first time, to utilize the full power of its graphics processors without any restriction to accelerate our hybrid, GPU/CPU-based, photorealistic rendering software, mental ray," said Rolf Herken, ceo/cto of mental images. "With the simultaneous introduction of its SLI technology, NVIDIA is taking the first step towards truly scalable GPU performance in commodity component-based 64-bit systems. Scaling the graphics performance by just adding another NVIDIA Quadro graphics board will be a very attractive option for our users."

"With this level of read-back performance and floating-point precision, any industry that requires the highest quality imagery can look forward to seeing a whole variety of previously CPU-bound applications accelerated in the back room by GPU," boasted Beth Loughney, gm of the NVIDIA Digital Film Group.