High-performance computing sounds so…futuristic.
But as the buzz around SC17 reveals, high-performance computing (HPC) is very much part of the daily workflow of most research universities, government agencies, and private sector organizations. With HPC fueling our journey to full embracing a data-driven future, we caught up with Doug Johnson, Chief Systems Architect and HPC Systems Group Manager at Ohio Supercomputer Center (OSC) to find out what the latest developments are in the field and what he’s looking forward to at SC17.
With clients from 26 colleges and universities in Ohio, and 41 private sector entities, Doug Johnson and his team are responsible for ensuring that these organizations have access to the most reliable and high performance computing environment. Since its inception in 1987 as a State of Ohio initiative, OSC’s mission has been to deliver a competitive advantage to its stakeholders. “Shared central HPC resources are more cost effective than the schools creating their own local resources,” said Johnson in a recent interview.
“By creating a shared environment like OSC, universities like Ohio State and Oberlin College are able to take advantage of the facility to help researchers solve some of the most pressing challenges we face today, not just in Ohio, but globally,” Johnson shared. From creating the next-generation of antibiotics that will provide doctors with new drugs to fight antibiotic resistant bacteria, to finding a solution to the toxic algae blooms that threaten the biodiversity and survival of the Great Lakes, OSC’s high-performance computing environment is in constant demand by researchers.
With increasing demands on the facility Johnson and his team have been doing some research of their own to determine how best to ensure that their compute environment supports researchers with availability and performance. “Researchers can’t afford downtime or data loss when they’re engaged in computations and simulations on which potentially years of research hinges,” Johnson shared. To avoid these pitfalls, Johnson has invested in state of the art storage arrays that reduce compute times, and ensure high availability for researchers, with an architecture that is resilient with no single point of failure.
It is these types of innovations that Johnson is eager to share at SC17, as well as learning more about how others are building their HPC environments to cope with increasing demand. “HPC is a vital part of our future and an essential tool for solving our most pressing societal challenges. While predicting what we will face in the future is hard, we know that design, simulation, and analysis using HPC resources will be important tools for researchers. Emerging tools take advantage of the ability to access large data sets, and use machine learning and deep learning algorithms to solve problems not available even a few years ago,” Johnson concluded.