STAMPEDE: Supercomputing for Big Data, Big Science
Supercomputing is for solving BIG problems, and doing so faster than 'regular' collections of servers. Supercomputing is fundamental for scientific advances, and is increasingly being used by companies industries to produce better products and to analyze exponentially increasing quantities of 'big data.' Every one of us benefits from supercomputing today--and will benefit even more tomorrow--and yet most of us do not realize it, or even know what supercomputing is.
This presentation will explain what supercomputing is, and why it is important (and cool!). It will present the design of the new, very-cool, world-class Stampede supercomputing system deployed by the Texas Advanced Computing Center, which offers several innovative capabilities including the first large-scale deployment of Intel's new Xeon Phi technology in the world. Finally, it will discuss how programmers can use Stampede to address big simulation and big data problems, and show some early interesting results.