The new GPUs deliver powerful enhancements including:

• up to 4 times the performance over previous GeForce generations
• innovative anti-aliasing technology
• nView(TM) multi-display technology

While NVIDIA has focused on the video game market, their products have always performed well for OpenGL-based MCAD operations. The new cards should be strong performers for pre- and post-processing CFD tasks.

CFX-ProMixus provides detailed analysis of mixing vessel processes through an embedded and fully automated computational fluid dynamics (CFD) solver. Using an intuitive interface, the user specifies reactor configurations from an extendable library of tank, baffle and impeller templates. CFX-ProMixus then automatically performs a CFD simulation of the flow within that vessel. The output is presented as a gallery of images and tables that permits pre-selection of component designs and reduces the number of iterations needed to develop optimum mixing performance.

CFX-ProMixus 2.1 adds many new impeller libraries, including new generic impellers (flat blade, pitched blade, retreat curve, curved blade and Smith turbine) and impellers from Philadelphia Mixers (high-solidity and low-shear hydrofoils), Chemineer, Cleveland Eastern Mixers and Lightnin (A310 and A320).

CFD 2002 is their 10th annual conference celebrating a decade of CFD excellence, is hosted by the University of Windsor and will be held at the Cleary International Centre in downtown Windsor.

The conference attracts academics, developers, and CFD users from industries such as Aerospace, Automotive, Chemical Process, Electronics, Energy, Manufacturing, and Power Generation. Anyone with an interest in exploring the latest advances in software and hardware is encouraged to attend.

Visit the CFD 2002 conference website at http://www.cfd2002.uwindsor.ca/eindex.htm for further information.

The Wildcat III 6110 and 6210 differ only in their onboard memory configurations. The 6110 features 64 MB of frame buffer memory, 128 MB of texture memory, and 16 MB of DirectBurst memory for a total of 208 MB of onboard memory. The 6210 features double the memory in each category for a total of 416 MB of onboard memory.

Other new features in the Wildcat III series are:

• Support for OpenGL 1.3 and OpenML 1.0
• Support for DirectX 7.0
• Support for multitexturing, bump-mapping and cube-mapping
• Up to 32 lights in hardware (versus 24 in previous versions)
• Dual DVI support
• Multiview and genlock support
• SuperScene antialiasing and 3D volumetric texturing

PC computing architecture has undergone evolutions to 16-bit (the 80286) and 32-bit (the 80386) computing in the past. Each step has provided better performance and more functionality for the PC platform. The step up to 64-bit computing will offer new functionality, but one question remains -- Is it worth it?

For some applications, such as large servers and specialized computing, a true 64-bit platform is necessary. However, many applications perform well on a 32-bit platform and avoid the memory bandwidth problem which comes from passing 64-bit words around.

Confusing the situation further is the fact that Intel and AMD are taking significantly different routes to 64-bit computing. Intel's Itanium processor is a completely new design -- one which was designed to run 64-bit code from the start. A side effect of this decision is that the Itanium executes 32-bit code slower than today's fastest 32-bit processors. While this may sound reasonable, recall that with the jump to 16-bit processing with the 286 and with the jump to 32-bit processing with the 386, the new processors still ran the old code base (8- and 16-bit, respectively) faster than the previous generation of chips.

AMD's new Hammer processors, however, are essentially extensions of their 32-bit processors to handle 64-bit operations. The result is a 64-bit CPU which will execute 32-bit code extremely well - a strong consideration given that 32-bit software will be around for a long time.

The air flow distribution in a spin coater has a critical effect on film thickness uniformity. The photoresist liquid contains a volatile solvent that evaporates during the spin coating process, and the final film thickness uniformity is determined by material spin-off and solvent evaporation. The solvent evaporation rate depends on the difference in partial pressures between the solvent in the air boundary layer next to the film surface and the bulk air over the surface, and on convective transport of air flowing over the wafer. If air flow near the wafer surface is different from the ideal flow induced by an infinite spinning disk, then the mass transfer coefficient at the wafer surface will change with radial position and the resist film thickness will not be radially uniform.