Over the last week, we managed to get some of the nation's biggest and baddest supercomputers to take a moment away from their gigabusy schedules and tell us what they were working on. They were happy to share.
Now: a word with iForge, the pride of the National Center for Supercomputing Applications.
Name: iForge
TOP500 Rank: Not Available
Vital Stats: System: a mix of 121 Dell servers, including 116 PowerEdge M610 servers featuring dual-socket/six-core (12 cores total) 3.46 GHz Intel Xeon x5690 processors, 96GB of DDR3 memory, and 292GB of storage; Three PowerEdge M910 servers featuring quad-socket/eight-core (32 cores total) 1.86 GHz Intel Xeon L7555 processors, 192GB of DDR3 memory, and 292GB of storage; and PowerEdge C6145 servers, which actually combine two complete servers in a single chassis. Each of these four nodes features quad-socket/12-core (48 cores total) 2.5 GHz AMD Opteron 6180SE processors, 256GB of DDR3 memory, and 600 GB of storage.
In English, that’s a 22-teraflop high-performance computing cluster that University of Illinois’ National Center for Supercomputing Applications (NCSA) farms out private sector partners, like Rolls Royce, Boeing, and Caterpillar.
What’s Happening: Supercomputers aren’t all about supernova simulations and super-hot plasma interactions, iForge will have you know. Roughly half of the supercomputers in the TOP500 work with private industry helping to make cars safer, materials more durable, or things like the Boeing 787 Dreamliner inexplicably stay aloft. Today iForge is in the employ of Rolls Royce, which you probably know as an automobile maker but should also recognize as a maker of all kinds of power systems, including jet aircraft engines.
Rolls Royce is running more than 50 supercomputing jobs, most of them lasting several hours and some rolling over into the next day. Most modeled complex fluid flow though aircraft engine designs, simulations that will help Rolls Royce design more powerful and more efficient engines. That in turn will make aircraft more reliable and less wasteful. That’s good for everybody. And it’s particularly great for Rolls Royce, which now doesn’t have to prototype every engine concept to see how it will stand up to real world conditions.iForge also notes that the NCSA is part of XSEDE, the Extreme Science and Engineering Discovery Environment, a collection of integrated supercomputing resources that helps find the right computers for the right jobs so that all of the computers in the network are working toward optimum output at all times. Other members of this network of supercomputing resources include Oak Ridge National Labs and the Texas Advanced Computing Center, from whom we’ve already heard this week. But don’t try to think about how powerful such a network of interconnected supercomputers really is, iForge warns. Unless you are actually a supercomputer, it will blow your mind.
Catch up with more supercomputers here.
Five amazing, clean technologies that will set us free, in this month's energy-focused issue. Also: how to build a better bomb detector, the robotic toys that are raising your children, a human catapult, the world's smallest arcade, and much more.


Online Content Director: Suzanne LaBarre | Email
Senior Editor: Paul Adams | Email
Associate Editor: Dan Nosowitz | Email
Assistant Editor: Colin Lecher | Email
Assistant Editor: Rose Pastore | Email
Contributing Writers:
Rebecca Boyle | Email
Kelsey D. Atherton | Email
Francie Diep | Email
Shaunacy Ferro | Email
Small correction: Jaguar is not part of XSEDE. Jaguar is one of the DoE's two leadership computing platforms, and XSEDE is an NSF project.
However, Kraken (another large Cray system located at ORNL but operated by the University of Tennessee, #11 on the last Top 500 list) *is* part of XSEDE; in fact, it is currently the most powerful system in the XSEDE project.
(Disclaimer: I'm one of the system administrators for Kraken.)
Well it SHOULD be all about supernova simulations and super-hot plasma interactions
-Knock knock
-Who's there?
-The Doctor.
-Doctor Who?
-Yes
Where I work at we have about 20 times the power of that used for a very mundane process everyday.
Pretty amazing how such an expensive device can save money. We replace people but in most companies they replace expensive testing. My Dad was a tool and die maker for some 45 years. Some of his work would be to make 10 bumpers for a future car mode. Then the car company would take those ten and find the very cheapest that would suit their requirements. They'd guess then test. The cost associated with all that build up and tooling had to be re-turned in the cost of a few million cars. Computer models can even save a car company from crashing 20 or 30 models into a wall. Very interesting stuff if they get the GIGO right.