From Science Node: “How supercomputers are uniting the US and China”

Science Node bloc
Science Node

12 July 2017
Tristan Fitzpatrick

38 years ago, US President Jimmy Carter and China Vice Premier Deng Xiaoping signed the US – China Agreement on Cooperation in Science and Technology, outlining broad opportunities to promote science and technology research.

Since then the two nations have worked together on a variety of projects, including energy and climate research. Now, however, there is another goal that each country is working towards: The pursuit of exascale computing.

At the PEARC17 conference in New Orleans, Louisiana, representatives from the high-performance computing communities in the US and China participated in the first international workshop on American and Chinese collaborations in experience and best practice in supercomputing.

Both countries face the same challenges implementing and managing HPC resources across a large nation-state. The hardware and software technologies are rapidly evolving, the user base is ever-expanding, and the technical requirements for maintaining these large and fast machines is accelerating.

It would be a major coup for either country’s scientific prowess if exascale computing could be reached, as it’s believed to be the order of processing for the human brain at the neural level. Initiatives like the Human Brain Project consider it to be a hallmark to advance computational power.

“It’s less like an arms race between the two countries to see who gets there first and more like the Olympics,” says Dan Stanzione, executive director at the Texas Advanced Computing Center (TACC).

TACC Maverick HP NVIDIA supercomputer

TACC Lonestar Cray XC40 supercomputer

Dell Poweredge U Texas Austin Stampede Supercomputer. Texas Advanced Computer Center 9.6 PF

TACC HPE Apollo 8000 Hikari supercomputer

TACC Maverick HP NVIDIA supercomputer

“We’d like to win and get the gold medal but hearing what China is doing with exascale research is going to help us get closer to this goal.”

___________________________________________________________________

Exascale refers to computing systems that can perform a billion billion calculations per second — at least 50 times faster than the fastest supercomputers in the US.

___________________________________________________________________

Despite the bona fides that would be awarded to whomever achieves the milestone first, TACC data mining and statistics group manager Weijia Xu stresses that collaboration is a greater motivator for both the US and China than just a race to see who gets there first.

“I don’t think it’s really a competition,” Xu says. “It’s more of a common goal we all want to reach eventually. How you reach the goal is not exactly clear to everyone yet. Furthermore, there are many challenges ahead, such as how systems can be optimized for various applications.”

The computational resources at China’s disposal could make it a great ally in the pursuit of exascale power. As of June 2017, China has the two fastest supercomputers in the top 500 supercomputers list, followed by five entries from the United States in the top ten.

1
Chinese Sunway ThaihuLight supercomputer, currently #1 on the TOP500 list of supercomputers.

“While China has the top supercomputer in the world, China and the US probably have about fifty percent each of those top 500 machines besides the European countries,” says Si Liu, HPC software tools researcher at TACC. “We really believe if we have some collaboration between the US and China, we could do some great projects together and benefit the whole HPC community.”

Besides pursuing the elusive exascale goal, Stanzione says the workshop opened up other ideas for how to improve the overall performance of HPC efforts in both nations. Co-located participants spoke on topics ranging from in situ simulations, artificial intelligence, and deep learning, among others.

“We also ask questions like how do we run HPC systems, what do we run on them, and how it’s going to change in the next few years,” Stanzione says.“It’s a great time to get together and talk about details of processors, speeds, and feeds.”

See the full article here .

Please help promote STEM in your local schools.
STEM Icon

Stem Education Coalition

Science Node is an international weekly online publication that covers distributed computing and the research it enables.

“We report on all aspects of distributed computing technology, such as grids and clouds. We also regularly feature articles on distributed computing-enabled research in a large variety of disciplines, including physics, biology, sociology, earth sciences, archaeology, medicine, disaster management, crime, and art. (Note that we do not cover stories that are purely about commercial technology.)

In its current incarnation, Science Node is also an online destination where you can host a profile and blog, and find and disseminate announcements and information about events, deadlines, and jobs. In the near future it will also be a place where you can network with colleagues.

You can read Science Node via our homepage, RSS, or email. For the complete iSGTW experience, sign up for an account or log in with OpenID and manage your email subscription from your account preferences. If you do not wish to access the website’s features, you can just subscribe to the weekly email.”

Advertisements