Titan supercomputer: ORNL launches giant computer capable of doing 20,000 trillion calculations per secondBy Angel Cuala on Oct 29, 2012 in Technology, United States •
The Oak Ridge National Laboratory (ORNL) launched on Sunday, October 28, 2012 its Titan supercomputer, a giant computer that is capable of doing 20,000 trillion calculations per second. The project, which is in Tennessee, is being supported by the US Department of Energy’s (DOE).
According to pdf file press release at OLCF.ORNL.gov that day, the Titan supercomputer will employ a family of processors called graphic processing units (GPU) which was first created for computer games, which can perform 20 petaflops, or over 20,000 trillion calculations per second.
As noted in the report, Titan will be 10 times more powerful than its predecessor, Jaguar, ORNL‘s earlier super high-performance level computer. Both devices have the same size, which is roughly the size of a basketball court, while each stack is about the size of a household kitchen refrigerator.
Titan is a Cray-XK7 system, which has more than 700 terabytes of memory, contains 18,688 nodes. Each of these nodes holds a 16-core AMD Opteron 6274 processor and a Tesla K20, NVIDIA‘s latest GPU accelerator. It will rely on its 299,008 CPU cores to guide simulations, allowing Tesla K20 to do its heavy task.
“Titan will allow scientists to simulate physical systems more realistically and in far greater detail.” James Hack, director of ORNL‘s National Center for Computational Sciences, was quoted in the report, with the supercomputer to be used computing power for research in energy, climate change, among others.
“The improvements in simulation fidelity will accelerate progress in a wide range of research areas such as alternative energy and energy efficiency, the identification and development of novel and useful materials, and the opportunity for more advanced climate projections.” Mr. Hack added.
“One challenge in supercomputers today is power consumption. Combining GPUs and CPUs in a single system requires less power than CPUs alone and is a responsible move toward lowering our carbon footprint.” Jeff Nichols, associate laboratory director for computing and computational science, said.
Image Credit: OLCF.ORNL.gov
Spread The News!