High Performance Computing (HPC) deploys many computers working closely together to solve computationally complex problems that are too large or would take too long for standalone desktop or server computers. Today’s HPC systems range from 100s to 100s of thousands of computers working in concert.
It’s about the computing…
The “Branscomb pyramid” shows different classes of computers. While attention is often focused on the pyramid’s apex – presently the realm of systems that operate in the PetaFLOP range – HPC at the large and medium scale is the focus of almost all academic and commercial R&D, where the lion’s share of results are being found and new discoveries made.
But even more about the applications and impact…
While it’s easy to get caught up in how much computing capacity a single facility has, “the purpose of computational science infrastructure is scientific discovery, not big iron bragging rights”. D. Reed, Corporate VP for Technology Policy, Microsoft.
HPC applications areas span most of science, engineering, and social sciences
including energy (e.g. fuel cells, underground gas/oil location) smart grid analytics, materials (e.g. high temperature superconductors), astronomy (origins of the universe), financial modeling, life sciences, industry product design, and much more.