6 June 2016 www.HPCSource.com
and now you see these levels of innovation, which are important for
a company of any scale to get ahead of its competition. Now, more
organizations are able to leverage high performance computing to
unlock their own innovations.
Q: Why is the continual advancement of high
performance computing so important?
A: High performance computing is so exciting to continue to watch
because of the applications where it gets deployed. There’s a relentless
pursuit for new levels of scalability and performance. We study the
dynamics in the ever-changing high performance computing market,
which is a fun segment to track because the industry relentlessly
chases new levels of performance and scalability. There’s always
going to be a new, more difficult problem to solve, a new discovery
to make, some invention to create. There’s no end to the number of
problems that our scientists and engineers are trying to solve, and
when they solve one, they’ll be off to the next one. High performance
computing is what enables this process, but you need to remain
continually focused on the industry’s future and that next, new
technology to empower these innovations on an ongoing basis.
About Addison Snell
Addison Snell is the chief executive officer of Intersect360 Research,
a market intelligence and consulting organization focused on high
performance computing. He launched the company in 2007, then
called Tabor Research, as a division of Tabor Communications, and
served as the company’s Vice President/General Manager. He and his
partner, Christopher Willard, Ph.D, acquired Tabor Research in 2009.
A veteran of the high performance computing industry, Addison
frequently shares his insights on forward-looking trends that shape
Addison was previously an HPC industry analyst for International
Data Corporation (IDC). Prior to IDC, he gained recognition as a
marketing leader and spokesperson for Silicon Graphics, Inc.’s
supercomputing products and strategy. Addison has a master’s degree
from the Kellogg School of Management at Northwestern University
and a bachelor’s degree from the University of Pennsylvania.
Q: People think of engineering, specifically in high
performance computing. Do you think it’s important
to focus on particular vertical solutions?
A: As we track the high performance computing market, the things
I get most excited about isn’t even the technology itself. It’s what
people do with the technology. And that’s important across industry
as well when you look at how has manufacturing improved, with
safer cars or self-driving cars or new materials that make things
lighter. How pharmaceuticals have improved with targeted drugs
that have fewer side effects, or energy where you can get more out of
an oil reserve without having to drill another hole. All of these are
important advancements in different kinds of industry and that goes
on in scientific research as well.
Probably what I’m most excited about is what’s going on in the
areas of brain research right now—think about when we mapped the
human genome. That sounds like the end of a problem—we did it—
it’s mapped, it’s done. The truth is, in science, that starts a whole
new field of research that created what genomics and brain research
will be like, where we get to a point where we have a comprehensive
map of the human brain. In a sense that’s the end of a problem but
in a much greater sense that gives you a whole new field of applying
high performance computing study of the neuropathy of brain
disorders from autism to Alzheimer’s or studying brain traumas. It
really is a wonderful tool that you might be able to use simulation
for because it’s more convenient than cutting open people’s heads.
It’s an important role for high performance computing
to continue to play out into the foreseeable future.
A big trend that we see going on in high performance computing
right now is the specialization of different technologies. We’re
coming out of an era where clusters of computers mostly looked the
same based on this notion of a Beowulf — and that was nice because
I could have my application and have it run on any platform without
worrying about changing it. But the new era of technology in high
performance computing is swinging us back more towards specialized
components, best-of-breed in different areas, and these different
processing elements in particular are going to require different
specialization in how you adapt the software to those solutions and
that’s where an industry focus on what is the application in question
— how am I tailoring this to a finance end user or seismic model or
something in life sciences? What’s the best configuration, the best
deployment for that solution going to be? It really requires a partner
who has a level of expertise across these different technologies in
order to maximize those innovations for end users.
Q: How is HPC becoming mainstream?
A: I think when a lot of people think of supercomputing the
automatic association is the really high-end research labs. Of course,
those are important to conduct research, but with high performance
computing, technologies have become more and more accessible by
a wider range of industry users across finance, manufacturing, life
sciences, energy, and chemical engineering.
These industry usages of high performance computing can
leverage industry-standard technologies that are bundled in
best-of-breed ways that are potentially accessible over the cloud,
Watch Addison Snell as he shares his thoughts on HPC
unlocking innovations for a competitive advantage.