July 6, 2016:
China recently announced it has built the fastest (t0 date) supercomputer (the fastest computers on the planet), one that is the first to achieve computational speeds of over 100 petaflops. A Peta is a billion billion and a flops is one floating point calculation per second. The Chinese TaihuLight hit 124.5 petaflops. What is more remarkable is that the TaihuLight is built from hardware designed and built in China. The record for computing speed changes fast and frequently but China is now established as a major player in this field.
While China stresses the many civilian applications of supercomputers the military has long been a major user of supercomputers. But use this was limited because of the size and cost of supercomputers. Thus supercomputers were too expensive for widespread use until graphic card maker Nvidia developed and sold graphic cards modified to be cheap and powerful supercomputers. In 2006 the Tesla supercomputer add-on for PCs appeared on the market. This was basically an Nvidia graphics board tweaked to act like a supercomputer, rather than a device that put 3-D, photo-realistic game graphics on your computer screen. By 2013 the latest version of these systems provided over five teraflops (a million billion) of computing power for under $7,000 (on one Quadro 6000 card, basically a tweaked graphics card). Using cards like this, some of the fastest and cheapest supercomputers in the world are built. At this point GPU (Graphic Processing Units, from high end graphic cards) based supercomputers were appearing on the list of 500 fastest supercomputers in the world. By 2016 most of the fastest supercomputers used at least some GPUs to achieve high speed and China has been an eager user of this technique. Not so much for TaihuLight but for the many jobs that require supercomputer capabilities, not one of the fastest ones in the world.
Supercomputers were first developed, as were the first computers, for military applications. These ultra-powerful computers are used for code breaking and to help design weapons (including nukes) and equipment (especially electronics). The military also needs a large amount of computing power for data mining (pulling useful information, about the enemy, from ever larger masses of text or visual information).
Because there's never enough money to buy all the supercomputers (which are super expensive) needed, military researchers have come up with ways to do it cheaper. Back in the 1990s it was American military researchers who figured out how to use GPUs for non-graphic computing. GPUs do something similar to what supercomputers do (lots of math calculations of a fairly simple type) and eventually the manufacturers of GPUs realized that there was a commercial (not just military) demand for GPUs serving as supercomputers.