The specifics of America’s upcoming exascale supercomputer were reported during a ceremony at Argonne National Laboratory where both Secretary of Energy Rick Perry and Democratic Senator Dick Durbin were in attendance. This new supercomputer from the Department of Energy is going to be built at the Argonne lab by Intel. And this machine will be the very first of its kind in the USA.
“We will use exascale computing and AI to accelerate discovery, spur ingenuity, drive innovation, and above all, we will impact all of those areas in ways that just a few years ago we couldn’t have realized that we were going to have the ability to do,” Perry mentioned.
This new supercomputer, which is called Aurora — which Secretary Perry mushed had been named after his black lab Aurora Pancake, who has only three legs — will be operational at the end of the year 2021. This is the DOE’s attempt to maintain America’s pace with China in the latest supercomputer technology race.
A February 2018 article in Science magazine had claimed that the two top Chinese supercomputers were actually more powerful than all DOE supercomputers combined – and there are 21 of these DOE supercomputers.
Summit, which is a machine that resides at the Oak Ridge National Lab in Tennessee, had actually reclaimed that title from the Chinese of being the most powerful of the world’s supercomputers when it went live in the summer of 2018.
It is believed that China will unveil its very first exascale supercomputer in 2020, which will surpass the United States once again in the supercomputer race. It is believed that Aurora — which has 50 times more power than Summit — will take back that title when it is fired up in 2021.
DOE Tries to Keep Pace with China in Supercomputer Global Race
“We don’t know what everybody else is doing, so we can only really talk to the plans in the United States,” says Rick Stevens, who is an associate lab director for computing and life sciences at Argonne. “We know other countries are working on the path to exascale, but we don’t know precisely when they will deploy their systems.”
Aurora had already been ushered in as a new standard supercomputer in the year 2015, but it got an upgrade to exascale in 2017. An exascale supercomputer is capable of operating at quintillion calculations per second. “When you talk about a billion billion [calculations] per second, for us mere mortals it kind of dawns on you where we are, what we’re doing, what we are on the cusp of being able to accomplish,” Perry pointed out. A combined project along with the National Nuclear Security Administration, this project’s goal is to boost America’s supercomputing power into the exascale realm by 2021, and Aurora will be the first such computer.
“So this system will be an excellent platform for both traditional high-speed computing applications” in addition to analyzing data, said Stevens. Aurora is going to be optimized for analyzing the various streaming data that come out of the DOE’s multitude of instruments; this includes telescopes, some particle accelerators, and several other detectors.
In accordance with current trends in today’s machine learning-enhanced technologies, Aurora will be established as the ideal platform for deeper learning. There are currently more than 100 artificial intelligence apps being developed in national laboratories alone, Stevens informed reporters.
“We kind of think of this as simulation data and machine learning as the targets for such a machine,” Stevens noted.
The Supercomputer will Help Identity Things like Suicide Risk Factors
Scientists that are presently working in various fields will likely discover that their projects will be enhanced by supercomputers. Material sciences — especially the constructing of more effective and efficient battery materials and other materials that are useful for alternative energies like wind, solar, and nuclear power — and even climate change forecasting projects. A combined project along with the Department of Veterans Affairs is going to set up the data-parsing abilities for identifying suicidal risk factors and intervention on the new supercomputer. This particular project will lead to numerous future leaps for humanity.
The potential defense apps for Aurora might also include the testing and analysis of strategic weapons since the US cannot test them physically anymore. If Aurora were to be assigned tasks like these, it could link up with additional DOE supercomputers to virtually test the entire nuclear arsenal of America.
A Contract that has an Estimated Value of $500 Million
Aurora is going to be constructed by the Intel Corporation along with Cray and Argonne, in a contract worth an estimated amount of $500 million. Rajeeb Hazra, who is a corporate VP from Intel and GM of the enterprise and government group, would not share the specifics of Aurora’s inner structure with the reporters who were in attendance. That included basic info like how much power Aurora would require — that data will be released at a later date. Hazra did tell reporters that they plan to work along with top research and academics in the nation to ensure Aurora will be equipped with the very latest bells and whistles after being booted up.
The DOE envisions these new exascale supercomputers as having a multiplier effect on new scientific discoveries like machine learning techniques and creating artificial intelligence. It is believed that these types of projects will become more commonplace.
“In general, we think this will create another wave of acceleration across many areas of science, technology, and health care,” stated Stevens.
Imagining a World with Supercomputers
Can you see a new world where artificial intelligence makes all of those pesky small decisions for you on a daily basis – the countless ones that are annoying? Perhaps you would review a report you are about to turn in at work while your self-driving car safely takes you to work. Your car already knows which route to take because it has reviewed current traffic conditions.
Things like crimes and traffic accidents could be predicted with amazing precision. Crimes that are committed are solved in a fraction of the time because of the incredible ability to review data and video footage.
Our learning would be online and greatly enhanced. Artificial intelligence would allow the machine to know if you have learned the information without testing you. And there would thousands of new occupations in the tech sector.
And this is only the beginning of what the future might look like, thanks to the first American exascale supercomputer.