Computer and Control Chapter (C&C) Dinner Meeting : vTools Events

IEEE.org | IEEE Xplore Digital Library | IEEE Standards | IEEE Spectrum | More Sites

Computer and Control Chapter (C&C) Dinner Meeting

#High #Performance #GPU #Computing #Graphics #cards. #GPU-accelerated #computing #is #the #use #of #a #graphics #processing #unit #(GPU).

High Performance GPU Computing By Justin McKennon, Senior Electrical Engineer NTS Lightning Technologies

In today’s world, it’s nearly impossible to imagine life without computers. They have become so delicately intertwined with every facet of our day to day lives. Computers have enabled us to solve problems that without them would be nearly unsolvable. However, as the complexity of these problems increase, so do the performance demands on computers. Until recently, it was commonplace for CPU manufacturers to simply add more transistors and increase the clock frequency of the CPU to increase performance. This approach however is not sustainable. We’re very nearly at the feasible limit for clock speeds, yet the never ending demands for increased performance exist. This has forced the computing industry to turn to alternative means to solve problems. This talk focuses on one of the most hotly researched areas in computers today: Graphics cards. GPU-accelerated computing is the use of a graphics processing unit (GPU). Since their inception, many of us have been wholly unaware that sitting inside this little card that makes your PC games run is actually a programmable supercomputer. If used properly, these can provide tremendous performance increases for a wide variety of problems.