Graphic Processing Unit
Over the past decade, graphics processors have made leaps and bounds and have been found to add significant value in both research and industry, where better performance in data intensive operations is needed.
Because of its structure, the GPU enables a single ‘instruction’ to be performed on huge chunks of data simultaneously (SIMD, Single Instruction, Multiple Data), compared to a general purpose CPU which typically has a smaller scale implementation of SIMD.
Think of the GPU as a coin press machine, which can punch out 100 coins with one operation from a single sheet of metal, whereas a CPU is a coin press which can punch out 10 coins at a time from a strip of metal. While the CPU might have a faster ‘time between punches’, it also requires a faster feed rate of metal strips as well. This is the key difference between the GPU and CPU. The GPU is throughput oriented, while the CPU islatency oriented.
The GPU is therefore well suited for operations that perform the same instruction on large amounts of data at once.