Wasting time = wasting life, Cuda prolonging your life
Once, the author of a web design work in the art friend told me a joke, their units for a large-scale promotional activities, ready to design a giant banner ads, and banner ads Design task is naturally their art responsibility. But it was not my friend who was specifically responsible for the design, but a few of his colleagues.
According to my friend's account, two colleagues, also designed for this huge banner ad, and equipped with the best computer for processing, but because of the large banner advertising size is too big, just a simple rendering process will wait half an hour or even the longest up to a morning time. Cause a friend's colleague to come down one day, can only do a few simple operations, and the rest of the thing is almost all sitting in front of the computer waiting ...
However, I do not know whether the reader can imagine that these long waiting, if using a GPU with powerful floating-point operations to carry out these operations, it will take only a few minutes or even a few seconds, but also to achieve the WYSIWYG form, which would greatly improve the working efficiency. If we were talking about a colleague of my friend who could have such a device, it's possible that a giant banner ad would take only a morning to complete.
CUDA, make your dream come true:
CUDA, the term is now more and more in our eyes. Although the cuda may not be so clearly related to our current computer applications, it is believed that with more and more software support in the future, Cuda will eventually go into each of our computer applications. The example we mentioned earlier, although it is not likely to be implemented at present, is likely to be seen in the next generation of Photoshop.
"With GPU acceleration, image rendering and processing will be greatly accelerated"
Support for GPU, physical acceleration on Photoshop Creative Suite next. Photoshop to open a 2GB, 442 million pixel image file will be very simple, zoom, rotate the picture without any delay. This is the embodiment of the GPU's powerful computing power, and to implement this function, it requires the help of an interface--cuda.
What is Cuda?