The last United States supercomputer to top the list of the world's fastest was Titan, in 2012. US President Donald Trump has taken an aggressive tone toward China, blocking the takeover of chipmaker Qualcomm Inc. on national security grounds and accusing China of stealing US trade secrets.
Because of Nvidia's GPUs, it makes sense that the computer's system will be used for machine learning and deep learning applications. The Summit supercomputer, which cost about $200 million, takes up the floor space of about two tennis courts and can deliver up to 200 quadrillion calculations per second-or 200 petaflops, reports Wired.
Summit operates at 200 petaflops while at maximum capacity. It is led by Jack Dongarra, a computer scientist at the University of Tennessee.
"The compute resources required by Summit and its workloads go well beyond how we would normally talk about flexibility and scalability for IT operations", Red Hat CTO Chris Wright wrote in a blog post.
"These are big data and artificial intelligence machines", said John E. Kelly, who oversees IBM Research.
If Summit performs as its handlers say it can, it should jump straight to the top slot of the Top 500 supercomputer ranking, which is calculated twice a year in June and November. The newest list will not be released until later this month, but Dongarra said he was certain that Summit was the fastest.
"Essentially, we are training computers to read documents and abstract information using large volumes of data", ORNL researcher Gina Tourassi said.
Supercomputing technology has been improving rapidly in recent years. It weighs almost as much as a commercial jetliner and is connected by 185 miles of fiber optic cables.
While impressive, Summit can be seen as a placeholder.
Supercomputers have advanced so far and so fast that it's easy to forget that the computers called smartphones we carry around in our pockets could stroll past a state-of-the-art supercomputer of a generation ago without breaking a digital sweat. And 95 percent of that computing power comes from GPUs. And China, Japan and Europe are developing machines that are even faster, which could mean the USA lead is short-lived.
From its start 75 years ago, ORNL has a history and culture of solving large and hard problems with national scope and impact, ORNL Director Thomas Zacharia said.
The eventual insights, Zacharia said, could "help us find new ways to treat our veterans and contribute to the whole area of precision medicine".