Virtual reality emphasizes visual experiences because vision gives the most information about the environment. The visual cortex is the most massive system in the human brain (Wikidoc, 2011), The eyes contain around 70 percent of all the body's sensory receptors, making sight the most information heavy of all the senses (Lefton & Brannon, 2002). Virtual reality generates visual signals that can be communicated as light rays to the user's eyes. Virtual reality images can be displayed on a computer screen or seen through special stereoscopic goggles. In 1968, Ivan Sutherland created the first virtual reality head mounted display system that gave the user the illusion of being in the same virtual environment as wireframe models (Pausch, 1997). Computer screens are less “real” than head mounted displays but much more popular due to higher resolution, lower cost, and ergonomics. The human brain is quick to perceive the 3D information implied by 2D images displayed on standard computer monitors. The video game industry has done much to ratify the idea that artificial worlds can be generated by software and adequately displayed on 2D screens. The movie industry also has used computer graphics to make movies since the 1970s. Toy Story in 1995 was the first movie made entirely with computer animation. Avatar in 2010 showed that computer animation could create a believable artificial world and that the visual illusion of 3D can be enhanced by special glasses. All of these methods can be used to visualize financial trading.
Graphics cards use dedicated GPUs to speedily render graphics on a local computer without bogging down the host CPU. Numerous exciting new graphics cards are introduced each year.
NVIDIA GeForce 400 Series® became available in 2010. Its Fermi® chip GPU has 512 stream processors in sixteen groups of 32 and 3.0 billion transistors. It supports OpenGL 4®, Direct3D 11®, VisualStudio® and C++. The card has 256MB attached to each of the enabled 6 GDDR5 memory controllers for a total of 1.5GB. The chip can perform one double-precision floating-point operation per cycle per core.
In 2006, NVIDIA® introduced CUDA®, a general purpose parallel computing architecture with a new parallel programming model and instruction set architecture. CUDA uses the parallel compute engine in NVIDIA® GPUs to solve many complex computational problems much faster than on a CPU. CUDA allows developers to use C or other languages such as Fortran, Python, Lua, etc. CUDA allowed the Fermi® card to be used for ray tracing by numerous applications shown at SIGGRAPH® 2010 in Los Angeles. Ray tracing is more realistic than shading so thus enhances the experience of virtual reality. The fact that the ray tracing can be done on the graphics card simplifies the software so that it will be easier to implement than on parallel computers. PEER1.com® uses the Fermi card to provide cloud supercomputer services. For example an auto design can be rendered in the cloud and the image displayed on an Apple iPhone®. Cloud supercomputing could be useful for traders who want to outsource supercomputing; when traveling; or in meetings when traders need a quick answer without breaking up to go turn on the supercomputer. Supercomputing often is used by virtual reality because of the difficult problems addressed.