broadsword
Brigadier
I know.
News on China's scientific and technological development.
- Thread starter Quickie
- Start date
I know.
Wuhan center to enhance country's supercomputing fortitude, acumen
By CHENG YU in Beijing and LIU KUN in Wuhan | China Daily | Updated: 2022-08-17 09:35
In a world where computing power may become the next strategic competition area, the Wuhan Supercomputing Center, China's largest container-like computing facility, was recently officially connected to the national computing network and is in its final launching stage.
Located in Central China's Hubei province, the landmark center will offer computing power of 50 petaflops in the initial stage, which is equivalent to 100,000 high-performance computers operating simultaneously, and is expected to eventually provide 200 pflops in total. Computing power roughly refers to the ability to process data.
The center has already been connected to the China Computing NET, or C2NET－a major national computing cluster that links supercomputing networks and artificial intelligence data centers across the country, including those in cities like Beijing, Xi'an, Nanjing and Qingdao.
"The move is a big step forward for the country's east-data-west-computing project. By leveraging local prowess in distributed computing, mass storage, AI computing power platforms and big data, the center will significantly drive computing power nationwide," said Feng Dan, a professor at the School of Computer Science and Technology, Huazhong University of Science and Technology.
The east-data-west-computing project involves China sending data gathered from the more prosperous eastern regions of the country to less-developed but resource-rich western regions for storage, calculation and feedback, which aims to help the country improve imbalances in the layout of digital infrastructure and maximize the value of data in a more productive way.
Although a late entrant to the high-performance computing field, China now tops the world with around 35 percent of the world's top 500 most powerful supercomputers and is quickly building a network of national supercomputing centers, where people can use computing power just like water and electricity.
The country's 14th Five-Year Plan (2021-25) also spelled out plans to construct 10 exascale computing systems by the end of 2025.
"China already surpassed the United States in AI computing power last year. I believe that in less than a year or two, the nation's overall computing power will exceed that of the US," said Wu Hequan, an academician at the Chinese Academy of Engineering.
"With more joint efforts in building supercomputer centers and networks, computing power will become the next new engine of the country's digital economy and further economic development," Wu said.
The Wuhan Supercomputing Center also adopted a new way of using container-like modular prefabricated boxes with integrated circuits and pipes laid in the box prior to being joined together, which shortens construction periods by one-third.
Wuhan center to enhance country's supercomputing fortitude, acumen
By CHENG YU in Beijing and LIU KUN in Wuhan | China Daily | Updated: 2022-08-17 09:35
The Wuhan Supercomputing Center, China's largest container-like computing facility, is close to completion in Wuhan, Hubei province. [Photo by WEI LAI/FOR CHINA DAILY]
In a world where computing power may become the next strategic competition area, the Wuhan Supercomputing Center, China's largest container-like computing facility, was recently officially connected to the national computing network and is in its final launching stage.
Located in Central China's Hubei province, the landmark center will offer computing power of 50 petaflops in the initial stage, which is equivalent to 100,000 high-performance computers operating simultaneously, and is expected to eventually provide 200 pflops in total. Computing power roughly refers to the ability to process data.
The center has already been connected to the China Computing NET, or C2NET－a major national computing cluster that links supercomputing networks and artificial intelligence data centers across the country, including those in cities like Beijing, Xi'an, Nanjing and Qingdao.
"The move is a big step forward for the country's east-data-west-computing project. By leveraging local prowess in distributed computing, mass storage, AI computing power platforms and big data, the center will significantly drive computing power nationwide," said Feng Dan, a professor at the School of Computer Science and Technology, Huazhong University of Science and Technology.
The east-data-west-computing project involves China sending data gathered from the more prosperous eastern regions of the country to less-developed but resource-rich western regions for storage, calculation and feedback, which aims to help the country improve imbalances in the layout of digital infrastructure and maximize the value of data in a more productive way.
Although a late entrant to the high-performance computing field, China now tops the world with around 35 percent of the world's top 500 most powerful supercomputers and is quickly building a network of national supercomputing centers, where people can use computing power just like water and electricity.
The country's 14th Five-Year Plan (2021-25) also spelled out plans to construct 10 exascale computing systems by the end of 2025.
"China already surpassed the United States in AI computing power last year. I believe that in less than a year or two, the nation's overall computing power will exceed that of the US," said Wu Hequan, an academician at the Chinese Academy of Engineering.
"With more joint efforts in building supercomputer centers and networks, computing power will become the next new engine of the country's digital economy and further economic development," Wu said.
The Wuhan Supercomputing Center also adopted a new way of using container-like modular prefabricated boxes with integrated circuits and pipes laid in the box prior to being joined together, which shortens construction periods by one-third.
What this ChinaDaily piece did not mention is that all these supercomputers-in-a-container are from Huawei.
This tweet is misleading in saying that the satellite "will be used for origin-tracing of infectious diseases, analyzing endemic diseases, etc."
In fact, the satellite will be used for detecting and tracking the "environmental factors (in water, earth and air) that can affect health of human beings".
The Japanese study is about top 10%, not top 1%.
All science-tech frontier advances and breakthrough come from Top 1%. The other important thing is the 'top lab'. Japan has atleast 8 World-class STEM r&d university labs i.e. 7 former imperial university, and Tokyo Institute of Technology. Korea has 7 i.e. Seoul National University, 3 engineering schools (KAIST+POSTECH+UNIST) and private SKY
(Sungkyunkwan+Korea+Yonsei). Though, Indian Institute of Science is on par with the weakest of the former Japanese imperial university i.e. Kyushu. There are no equivalent university lab in Italy or Spain. Best Italian university lab is Sapienza, which is roughly on par with Tsukuba or Keio in Japan, or Hanyang in Korea. Moreover, Japan has nearly insurmountable 3rd position in core patents.Now, real world is defined by Qimen Dunjia
i.e. the 8 foundation technology i.e.- AI (including robotics which itself includes autonomous vehicle, drone, computer-integrated machines)
- Quantum
- Chip (end-to-end chip industry)
- Communication (electrical wireless like 5g/6g, Satellites, and Photon)
- Brain (including Neuromorphic Computing)
- Biotech
- Energy (nuclear including high-power laser/Peta & Exa & beyond, solar, molecule like battery & hydrogen)
- Space
Incorrect. The Japanese study tracks both measures.
From SenWenYu's previous comment:
A critical aspect of China-US competition will be STEM talent. I don't think most people - including most Americans - understand just how dependent the US is on foreign talent to fuel its tech ecosystem.
China has surpassed the US in STEM graduates years ago - and the gap is growing.
International co-operation is good for China, but losing talent isn't. The good news is that fewer Chinese top students go abroad (and even among those who go, many come back), this will be a major problem for the US going forward. China should make studying abroad harder, especially as Chinese universities are becoming better and better.
Just look at how Tsinghua and PKU have improved in the past few years.
This trend of improvement is similar for other Chinese universities. It should help alleviate emigration pressure of the brightest talents.
China has surpassed the US in STEM graduates years ago - and the gap is growing.
International co-operation is good for China, but losing talent isn't. The good news is that fewer Chinese top students go abroad (and even among those who go, many come back), this will be a major problem for the US going forward. China should make studying abroad harder, especially as Chinese universities are becoming better and better.
Just look at how Tsinghua and PKU have improved in the past few years.
This trend of improvement is similar for other Chinese universities. It should help alleviate emigration pressure of the brightest talents.
Last edited:
It'd be interesting to look at the more recent data. It's quite likely to have changed dramatically, considering the # of Chinese young adults that are still looking to study in the US has already dropped significantly.
Additionally, I'd be extra careful at quoting any Western based institutions' rankings, especially a British for profit at that. The amount of small minded prejudice so prevalent in Western societies (the Anglo Saxons being the worst offenders and propagators) almost certainly would have seeped through into any "rankings" or "surveys" that are really just aimed to reinforce their own beliefs which are quite detached from reality.
The photo you uploaded titled "Highly cited scientific papers country ranking" is about top 10%.
About top 1% papers, Other than China & America, both around ~4500 papers, rest are statistically too diminutive, and hence quality of top labs making frontier advances and breakthrough is more important metric.
As far as QS rankings, it's nothing but trash. The closest ranking to scientific quality is Nature Index [Only Physical & Biological Sciences], even then Nature Index has many loopholes like
- "Earth and Environmental Sciences" instead of top journals from foundation fields like Optica, Optics Letters, Optics Express, Nuclear Fusion [IAEA] ..
- Journals of European Molecular Biology Lab and European Southern Observatory, Journal of German Chemical Society, Journal of Cold Spring Lab, Journal of Rockefeller University, and European Physical Journal.
- Absence of top Chinese Journals where huge chunk of China's premier works are published i.e. Chinese Physics Letters, Photonic Research [by Changchun Institute of Optics of CAS], High-Power Lasers [by Shanghai Institute of Optics of CAS and National University of Defense Technology] and Plasma Science and Technology [by Institute of Plasma Physics of CAS].
- Foundation science-tech frontier advances and breakthrough
- Nature Index [Physical & Biological Sciences]
- Prestigious AI conferences, Machine Learning [NIPS+ICML+ICLR], Vision [ICCV+CVPR], NLP [ACL+EMNLP], Robotics [ICRA] and Applied AI [AAAI+IJCAI].
- Engineering Scale.