The glory of having the world’s fastest supercomputer, as measured by the Linpack benchmark, has been China’s for four years running, first with the 33-petaflops Tianhe-2 and currently with the 93-petaflops TaihuLight. The country also set an ambitious goal to reach exascale by 2020 and is funding at least three competing exascale prototypes.
News out of China today indicates the nation could stand up a peak exaflop supercomputer even sooner. The South China Morning Post reported that China will build its first exascale machine on the Shandong province coast as soon as 2019 to support ocean research in the South China Sea and boost China’s maritime expansion.
“The most important question to us is not whether China can build an exascale computer, or how fast, but why,” An Hong, professor of computer science with the University of Science and Technology of China in Hefei, told the South China Morning Post. “There is indeed a race among nations on supercomputers, but this is not our concern. Our concern is the ocean.”
There are three supercomputing makers competing for the contract: Sugon, formerly known as Dawning Information Industry, which is owned by the Chinese Academy of Sciences; the National University of Defense Technology (NUDT), which designed the Tianhe supercomputers, and the Sunway team that built TaihuLight.
Ocean simulation was one of the workloads at the ASC17 student contest at the Wuxi Supercomputing Center (home of TaihuLight) in April. The 20 student team finalists were tasked to complete a deep parallel optimization of the high-resolution maritime data simulation code MASNUM on TaihuLight.
The exact site for China’s first exascale-class supercomputer has not been decided, but people familiar with the project reason that the city of Qingdao, the largest port in Shandong, is a likely candidate. The world’s largest marine datacenter is being built there and it has direct links to the major ocean monitoring networks.
The project reflects the mission of Chinese Communist Party leader President Xi Jinping to turn China into a “hai shang qiang guo,” or maritime superpower.
“It will help, for instance, the simulation of the oceans on our planet with unprecedented resolution. The higher the resolution, the more reliable the forecast on important issues such as El Nino and climate change,” Feng Liqiang, operational director of the Marine Science Data Centre in Qingdao, Shandong, told the South China Morning Post. “It will give China a bigger say in international affairs.”
Not everyone agrees that the project will catapult China to maritime preeminence. Chinese ocean scientists pointed out that the US holds significant advantages in terms of having decades of historical data, sophisticated HPC software and algorithms, as well as a culture of open access to scientific data.
“The State Ocean Administration runs and hoards its own data sets, as do the PLA Navy, the Chinese Academy of Sciences and many universities. Every institute is treating data as private asset for the interests of their own research,” Lu Xianqing, professor at the Key Laboratory of Physical Oceanography under the Ministry of Education in Qingdao, told the South China Morning Post.
The pricetag for the project is estimated at one to two billion yuan (US$150 million-US$300 million).