While America appears to be creating supercomputers with the sole intention of making them stars of daytime TV, China seems to be taking the task altogether more seriously.
After it officially laid claim to the most powerful supercomputer in the world with the Tianhe-1A, China will now be looking to continue the development of high speed computers with chips made on its own turf, another step in the one-upmanship with the United States.
Now, according to scientist Hu Weiwu, Chinese Longsoon chips will be used as the core component on the forthcoming Dawning 6000 supercomputer, with a computing speed of 1,000 trillion operations a second.
“Our information industry was using foreign technologies. Nonetheless, just like a country’s industry can not often depend on foreign steel and oil, China’s information industry requires its own CPU,” Hu said.
Making supercomputers is big business in China, with the Institute of Computing Technologies of CAS, Jiangnan Institute of Computing Technologies along with the National University of Defense Technologies (NUDT) all conducting projects which will solely use domestic chips.
While the Dawning 6000 will not reach the stellar speeds of the Tianhe-1A, which required 14,336 Intel CPUs and 7,186 GPUs from Nvidia, it is expected to focus on being more energy efficient, using 10,000 Longsoon microchips, according to Chinese media.
“Each year the electricity bill could price much more than 10 million yuan ($1.5 million) for one supercomputer, and we are only using one tenth of its capacity at most,” Hu said.
However Hu has admitted that the chips themselves have a long way to go if they are going to stand toe-to-toe with Intel’s.
“It still requires an additional decade just before China-made chips meet the requirements of the domestic marketplace. Hopefully following two decades, we will have the ability to sell our China-made CPUs to the US just like we are selling clothes and shoes,” Hu said.