Graphics processing units (GPUs) have long been a means to shift graphic-intensive computation away from the CPU. GPUs are designed to be parallel having hundreds of cores vs. 2-4 cores in traditional CPUs. Software developers are increasingly looking to GPUs for non-graphical computational-heavy operations to achieve improvements in efficiency and power consumption. This is known as general purpose computing on GPUs, called GPGPU. The challenge is learning how to program systems that effectively use these concurrent processors to achieve efficiency and performance goals. Programming professionals and advanced computer architecture and programming students need to know how to program these processors. GPU Computing Gems includes tested, proven GPGPU and CUDA techniques from the leading minds of concurrent programming, offering insights unavailable in any one volume to date. The contents cover the breadth of industry from scientific computing and engineering to artificial intelligence, with techniques including statistical and financial modeling, rendering, computer-aided design, and more.