Page 1 of 2
Nvidia on Monday expanded its Quadro line of graphics processing units (GPUs) based on its Fermi architecture to include the Quadro 2000 and Quadro 600.
The Quadro 600 graphics card is for basic, entry-level users and features 96 CUDA processor cores, while the more powerful Quadro 2000 runs on 192 CUDA cores. The Quadro 600 is only half the height of the 2000 model but it can interact with larger more complex models and offers more efficient performance per watt.
Both Quadro cards run on CUDA (Compute Unified Design Architecture), a parallel computing architecture available to software developers for a number of applications. CUDA's merits as support for improved visualization were on display at last month's GPU Technology Conference in San Jose.
"Parallelism on a CPU is task-based, it only scales up and there is only so many things that can run at the same time," an Nvidia spokesperson at GTC told CRN. "Instead we have chosen a data-parallel approach to create for flexibility in order to break through the limitations of the parallel power-scaling brick wall."
In addition to Fermi and CUDA, Nvidia is now offering a Scalable Geometry Engine Technology for computer-aided design applications testing virtual models of various products. This offering joins the updated Quadro line, a workstation graphics solution for Digital Content Creation applications including SolidWorks and Autodesk 3ds Max.
Next: How Quadro Fits In With Nvidia Technology
