HISTORICALLY, THE UNDERSTANDING of classical computer systems architectures has focused squarely on the interaction between the central processing unit (CPU) and the memory infrastructure. However, a new breed of system is upon us, in which the graphics processing unit (GPU) plays an integral role and is as important as both these key components.
As software developers and consumers have demanded increased photorealism from games and more complexity and fluidity from their user interfaces, the requirements of computer graphics have increased. The humble GPU has been catapulted from a simple line-drawing accelerator to a highly parallel, multithreaded subsystem in its own right, with such computing power that it has become integral to modern computer architectures.
However, to understand the potential of graphics technology we must focus on its primary purpose and make sense of it in the context of modern 3D graphics.
Although William Fetter is credited with coining the term “computer graphics” to describe his work on human body animation with Boeing in the early 1960s, the origin of 3D graphics can be traced back to the 1950s and military flight simulators (see Figure 10-1). As early as 1951, the Whirlwind computer at the Massachusetts Institute of Technology (MIT) was being used as a visualisation tool. The Whirlwind computer allowed oscilloscope-style graphics with user input via a device resembling a light pen. The Whirlwind ...