NEW YORK: Intel Corporation is planning to launch ‘Aurora’ world’s most powerful supercomputer to date, with peak processing power at a record 180 petaflops in 2018. It will be based upon the Cray’s next-generation supercomputer framework code-named Shasta. There is not much information known about the Shasta supercomputer, other than the fact that it will have a quicker interconnect than what Cray is currently offering.
It is imperative that a higher bandwidth network is put in place to reach 180 petaflops of performance to enable faster data transfers between processors, storage, and memory units. Shasta is created by tech firm Cray.
The Shasta computer has been chosen by the Department of Energy as the supercomputer of choice, in a $200 million agreement. The introduction of this new supercomputer is seen by analysts as a way for the US to gain edge over China in the supercomputer arms race, which is probably why the US is investing to create a 180-petaflop supercomputer that can be utilized for scientific research.
The Aurora will be deployed by around 2018. Meanwhile, there are other researches being conducted where systems will be able to churn out performances averaging between 150-300 petaflops.
The US gives high priority to supercomputers that it considers to be vital to the security, economic, scientific, and environmental programs in the country. Supercomputers are used in the military for defense purposes, weather forecasting, and for the creation of real-time economic simulations. It has become imperative for any country to boost its various departments by employing supercomputers.
At present, the fastest supercomputer in the world is the Tianhe-2 in China, which has the potential of giving a peak performance of 54.9 petaflops, according to Top500.org data. The Aurora system is stipulated to be a newer version or an upgrade to the Department of Energy’s Titan supercomputer, which is the second-fastest computer in the world.
After some time, supercomputers and their embedded technology are incorporated into normal servers, enhancing the overall computing power of an average PC. This is due to the tendency of higher technology flowing from supercomputers down to servers used in data centers, after a certain time lag.
Once it’s fully functional, the new Aurora supercomputer will be incorporated at the Argonne National Laboratory in Lemont, Illinois. Commenting on the development, Intel said that the supercomputer will incorporate “next generation” Intel Xeon chips having the capacity of up to 72 cores. Intel will also use silicon photonic technology to help transport data through supercomputers with pulses of light.
