Cray company has announced the launch of a new supercomputer Cray XC30 (codename Cascade), which will be able to set a new performance record for this class of systems. Recall recently launched Titan supercomputer, peak performance in excess of 20 petaflops (20 million billion floating point operations per second). However, the Cray XC30 Supercomputer can provide 5-fold increase in productivity with its estimated peak performance of 100 petaflops.

Cray XC30 Supercomputer partially developed as part of DARPA High Productivity Computing Systems. It uses topology Dragonfly. The system will use a family of Intel Xeon E5-2600, most in plans to integrate more than 1 million cores. In future versions of the XC30 developers intend to implement a co-processors Intel Xeon Phi and graphics NVIDIA Tesla, which will ensure the continued productivity growth. The new supercomputer will be used interconnect Aries and special software environment. In addition the implementation of new technologies in the field of power and cooling is done. In particular, there is the use of cooling systems with intersecting air currents. This is intended to reduce the amount of operating costs.

This Cray XC30 Supercomputer would be launched in first quarter of 2013, so stay tuned for more updates regarding this supercomputer.