Supercomputing apps migrate backwards

My pal Trey sent me an email yesterday with a pointer to an article in Computerworld about applications that are migrating backwards: going from supercomputers to desktops. This isn’t a new thing, but the advent of GPUs and CUDA has made it even more reasonable for non-heroic users to move to desktops if they can. There are good incentives for them to do this if they can still get their work done: no queues, no fighting with the big boys for allocation, and direct control over the resource.

Some of the interesting apps in the article

At Temple University, researchers have developed models that measure the effects of applying anesthesia on molecules within nerve cells. The models currently run on a supercomputer, but plans are underway to perform the calculations on an Nvidia GPU cluster with four nodes. This will both save money and give researchers more flexibility to conduct tests when they’re ready to do so (instead of having to wait for their scheduled time to use a supercomputer).

…Researchers at the Ohio Supercomputer Center (OSC) in Columbus, Ohio, have found that not every simulation requires a traditional supercomputer. Don Stredney, the director and interface lab research scientist for biomedical applications at OSC, found a limitation that’s common with supercomputers: Batch processes are static and run on a scheduled time frame. They cannot provide real-time interactions, so they can’t mimic a real surgical procedure. Desktop workstations that cost $6,000 to $10,000 allow his team to run simulations that show, in real-time, how a surgery changes a patient’s anatomy, he says.

…Injection-molding simulations are invaluable to car makers, Autodesk’s Martin says…Simulations “used to require a significant cluster-computing installation, but we are achieving the same level of power with current desktop computers,” says Martin, who says the desktop advances that played the biggest roles in making that possible were the move to multicore processing, the use of multiple GPUs and 64-bit throughput. Martin uses standard desktop computers that can be purchased at Wal-Mart, with the latest 3D-capable GPUs and Intel dual-core CPUs.

More interesting, complex applications in the article.