The Titan supercomputer was created out of the supercomputer Jaguar by adding Nvidia's Tesla GPUs to existing AMD Opterons. The system consists of a 200 cabinets containing 18,688 computational nodes. Each of them is running 16-core Opteron 6200 series and Nvidia's Tesla K20 GPU accelerators. The computational points are connected via Cray's high-performance Gemini interface that allows the Titan to increase performance over 50 petaflops with future upgrades.

In comparison to its predecessor Titan is more than 10 times faster and 5 times more energy efficient. The previous system named Jaguar occupied the same floor space as the Titan is now. The supercomputer will be used for research of biological and physical phenomena and to give results of certain experiments faster than it would take scientist to perform them manually. Jeff Nichols, associate laboratory director for computing and computational sciences at Oak Ridge National Laboratory said “Science and technology have always been our primary goal, and Titan is a ground breaking tool that will allow researchers worldwide to leverage GPU-accelerated computing to make unparalleled breakthroughs!”. There is no doubt this will help them achieve their goals, or at least to get closer to them, until an improved version of this supercomputer is constructed.

[Ed – it really was only a matter of time before someone decided to leverage the existing computational power of a super computer with nVidia’s Tesla GPGPUs. We are a little surprised that we are seeing an AMD/nVidia combination we had always expected to see nVidia and Intel team up with Cray for this spot, but still it will bring new compute power to the table and allow for existing systems to be upgraded more efficiently than previously possible. We can only dream of what might be next.]

Tell us what you think about this Titan in our Forum