I came across this interesting video while checking on those I follow on Twitter; this is really cool: [youtube]http://www.youtube.com/watch?v=1-tMp4WkQjA&feature=player_embedded[/youtube]
I’ve never been much into video games, but the technology can be leveraged for more interesting things as seen in the video above.
Innovation usually isn’t about anything brand new, but often re-using something designed for one purpose and re-applying it in another application. This could be re-use of technology, or even re-use of methods. After all, a very wise man said, “there is nothing new under the sun”.
For example, the Grahics Processing Unit (GPU) were designed to offload graphics computations from the Central Processing Unit (CPU) to accelerate graphics on computers and reserve the CPU for other tasks, improving overall performance. This technology was picked up by the gaming industry, which lead to games machines with amazing 3D graphics that could be rendered in real time. With the high volume of that market, these powerful chips became really cheap. Now GPUs are making there way into supercomputers. Not to offload graphics processing, but to offload complex math from the CPU, and when the parallel program is designed effectively for this architecture, orders of magnitude performance improvement can be achieved, for relatively cheaply. In fact, there are many in the High Performance Computing (HPC) industry who now question the Top500 scoring, as GPUs make it easier to build a machine with a theoretical world record performance. The top ranked computer is now from China, and the score is heavily impacted by GPUs built into that machine. Some argue that machines in the top 10 that do not use GPUs are in fact far more powerful, as they would perform much better on programs that are not optimized for GPU acceleration, thus more generally usable. But I digress.
When I first began working in the supercomputing industry, it was clear that there was much opportunity for re-use from the telecommunications industry. The telecomm business had long ago discovered that operational costs quickly exceeded capital costs. Looking at supercomputing architecture back then, at least some of the ‘new’ commodity cluster implementations, it did not appear that this reality had hit home to most managers. A simple example was power distribution. These systems simply used 120VAC. That’s fine for your refrigerator or PC, but for large power hungry systems there are far more efficient ways to distribute and protect power supply.
I’m sure the blog idol participants can think of more interesting examples of innovative technology re-use.
Not only that, we have never before had so much technology so easily available for anyone to innovate with. I believe we are at the cusp of a great wave of innovation, and those willing to think creatively about using commodity technologies in new ways, will be the next generation of great inventors.
And not only technology, with the trend toward open data, there are even more opportunities to be innovative in creating new services re-using existing data.