Since 1986 - Covering the Fastest Computers in the World and the People Who Run Them

Language Flags
April 14, 2010

Cloud Enabled: The New HPC Consumer

Nicole Hemsoth

Let me fill you in one a little secret only my word processing software knows.

The original title of this post was, “What Happens When Everyone Wants Their Very Own Supercomputer?”

But not only was a question like that setting up a parade of more questions, all of which were far too complex to proffer tentative answers for in an entire dedicated series, it simply missed a certain salient point from the get-go—every aspect of cloud computing in the HPC sphere, from infrastructure, middleware, software, and general interface or control—is becoming…well, accessible. And ome might (rightfully?) argue that HPC was never meant to be that way; you know, accessible?

We all know why this new, Model-T-like influx of widespread accessibility of the once-unimaginable exists though, don’t we? You guessed it–it is the usual suspect.

 Economics.

And if economics is the root source of innovation (let’s please not fight about that right now) then it’s clear why HPC is moving out of the exclusive domain of scientific and academic worlds and into the realm of enterprise and mid-sized commercial ventures.

Good old-fashioned consumer demand.

Because after all, this is a consumer base that is being targeted in grander ways by HPC and cloud computing giants like Amazon—just to name one that does a particularly nice job of complementing the giant imagery. And yes, you heard me right, I said “consumer” rather than research organization. Rather than academic institution. Rather than specialized data-driven enterprise.

Calling this new, growing population of HPC enthusiasts (and suspicious parties dipping one toe in to test the water) a “consumer base” is important in itself because it implies a certain mainstream type of consumption; and that is a type of consumer who we might have once been considered as having needs that were far too frivolous in the face of grand ultra-machines. But everyone wants a supercomputer–and it’s the job of the cloud to deliver that possibility to the masses, right?

The once incomprehensible possibility of harnessing the power of a supercomputer at a reasonable cost by migrating into the cloud is far too appealing to pass up. It follows then that these next few years will be a time of fierce innovation because demand exists—lots of it. And its coming from all quarters—it’s just that some of the quarters have never really needed or considered (or both) the possibility of high-performance, high-productivity computing to manage their multidimensional workloads.

Now that I think about it, perhaps a more compelling title to spur future topical exploration is the other end of the title’s question: “What Happens When Everyone Finally Has Their Very Own Supercomputer” … Something tells me, however, that the answer to that question is not long in coming as the innovations, trials, and errors of the next decade continue to play out. 

The once far-off notion of small business supercomputing is already here. The cloud delivered it.