Making Demands on On-Demand

You probably know this already because it has been hard to miss -- the current hot buzzword is "on-demand." Okay, technically that's two words, but when you hyphenate them.... Well, you get the idea.

Lately, we've heard about "on-demand security," "on-demand collaboration," on-demand CRM," "on-demand applications," "on-demand supply chain," "on-demand computing software environments," "on-demand strategies," and on and on. There's even an OnDemand Expo in the offing. There's Oracle On Demand, Adobe On Demand, and On-Demand IBM.

All this, even though Wikipedia says that "SaaS" has replaced "on-demand". You couldn't prove by me. I just received promotions for "on-demand adult videos" not to mention "on-demand router-to-router VPN connections" both in the same day.

Gee, if you didn't know better, you'd think we're an impatient lot.

But that's not to say that on-demand is unreasonably demanding. Say you were in a natural disaster -- like an earthquake -- and needed some help. To thisend the San Diego Supercomputing Center at the University of California-San Diego has introduced OnDemand, a new supercomputing resource that will support event-driven science.

The OnDemand system is a Dell cluster with 64 Intel dual-socket, dual-core compute nodes for a total of 256 processors. The 2.33 GHz, 4-way nodes have 8 GB of memory. The system, which has a nominal theoretical peak performance of 2.4 Tflops, is running the SDSC-developed Rocks open-source Linux cluster operation software and has the IBRIX parallel file system. Jobs are scheduled by the Sun Grid Engine. The system has StarP installed, a parallel version that provides a high-performance backend to packages such as Matlab.

The system is already in operation and formal allocations of time for the OnDemand system will begin in October. In addition to supporting important research now, this system will serve as a model to develop on-demand capabilities on additional TeraGrid systems in the future. TeraGrid is an NSF-funded computing grid linking some of the nation’s largest supercomputer centers including SDSC.

"SDSC's new OnDemand system is an important step forward for our event-driven earthquake science," said Caltech computational seismologist Jeroen Tromp. "We’re getting good performance that will let us cut the time to deliver earthquake movies from about 45 to 30 minutes or less, and every minute is important."

Posted by Jon Erickson at 09:28 AM  Permalink