Nvidia releases shedloads of cards and software

GRAPHICS CHIP DESIGNER Nvidia has announced a raft of software to support its Fermi GPUs and proclaimed a new era of graphics with its Fermi based Quadro line of workstation graphics cards.

First were some Application Acceleration Engines based on the Fermi architecture. These latest versions designed for the professional software development ommunity include Nvidia's Scenix 6, Cg Tooklit 3 and Optix 2.

Nvidia also demonstrated its Iray renderer in commercial applications as well as what the Green Goblin calls a breakthrough in GPU cloud-based rendering.

Jeff Brown, the general manager of Nvidia's professional solutions group said that the company's latest engine releases enable developers to bring software to market far faster.

"They open the door to a new generation of software which combines advanced visualisation with high performance computing and simulation," he said.

Scenix 6, Cg Toolkit 3, Optis 2 technology, and Iray rendering are supposed to give software developers new features and benefits that they can offer to end-users across a broad range of markets, including medical imaging, automotive styling, architectural design and energy exploration.

Nvidia's Scenix 6 has a new bezier patch geometry class using Cg tessellation with OptiX 2 support for faster interactive ray tracing and improvements in performance and fidelity.

Its Cg Toolkit 3 has new tessellation programs allowing displacement and procedural surfaces to dynamically adapt their tessellation in real-time on Fermi GPUs. There are also OpenGL 4 and DirectX 11 levels of programmability for the latest in portable, cross platform effects.

Optis 2 has optimisations for Fermi and Nvidia claims it has four times the performance over previous generation (GT200) GPUs and ten times over the G92 series. It has support for CUDA on Windows, Linux and Mac OSX, as well as for Direct3D and fast interoperability in Direct3D and OpenGL.

Nvidia Application Acceleration Engines are available for download immediately to developers.

At the same time Nvidia has declared that this is the age of the 'computational visualisation workstation'. Basically this means that we are talking about the Quadro workstation GPUs that Nvidia claims deliver performance that is up to five times faster for 3D applications and up to eight times faster for computational simulation,

Nvidia has also released its Quadro Plex 7000 array, Quadro 6000, Quadro 5000 and Quadro 4000 GPUs. These feature the new Scalable Geometry Engines and use the Application Acceleration Engines software.

The Quadro GPUs are built to the OpenGL 4.1, DirectX 11, Directcompute and OpenCL standards.

The Nvidia Quadro professional range includes the Quadro Plex 7000 with 12GB of GDDR5 memory and 896 CUDA cores, the Quadro 6000 with 6GB of GDDR5 memory and 448 CUDA cores, the Quadro 5000 with 2.5GB of GDDR5 memory and 352 CUDA cores, and the Quadro 4000 with 2GB of GDDR5 memory and 256 CUDA cores.

On the mobile workstation front there is the Quadro 5000M mobile workstation GPU with 2GB of GDDR5 memory and 320 CUDA cores.

All Quadro products are compatible with the Nvidia 3D Vision Pro active shutter-glasses.

As to pricing and availabilty, the Quadro 4000 (£779) and Quadro 5000 (£1,709) are available immediately. The Quadro 6000 (£3,579) will be available this autumn and the Quadro Plex 7000 will also be available this autumn with pricing yet to be announced. Mobile workstations based on the Quadro 5000M will be available in the third quarter from HP and Dell. µ