News on China's scientific and technological development.
- Thread starter Quickie
- Start date
Dedicated scientists help build supercomputer
By ZHANG ZHIHAO | CHINA DAILY | Updated: 2022-07-27 09:00
In the computer room of the National Supercomputing Center of Tianjin, rows of neatly packed black cabinets work on mankind's biggest scientific questions, ranging from the evolution of the universe to modeling complex protein structures for drug discovery.
The machine, called Tianhe-1, is named after the river of stars in the night sky that makes up the Milky Way. For centuries, Tianhe has been a source of inspiration and exploration in Chinese culture, but in Tianjin, it is now synonymous with innovation.
The instrument is China's first petascale supercomputer capable of averaging over 2.5 quadrillion operations per second. It can handle more than 1,400 computing tasks simultaneously and around 1,000 research groups use the supercomputer every day, said Meng Xiangfei, the creator of Tianhe.
Meng and his colleagues are currently building the prototype of China's new-generation exascale supercomputer called Tianhe-3 that can do over 1 quintillion operations per second, which will be orders of magnitude more powerful Tianhe-1.
About 15 years ago, Meng didn't believe China could independently develop a supercomputer, let alone become a global supercomputing powerhouse.
According to the latest edition of Top500 in June, a list of the world's fastest supercomputers, China has 173 such systems, followed by the United States at 128. China's current most powerful supercomputer is the Sunway TaihuLight, which is ranked sixth in the world.
Meng, 43, is now the chief scientist at the department of application and research at the National Supercomputing Center of Tianjin. Since joining the Communist Party of China around 20 years ago, Meng has dedicated himself to the development of China's supercomputers.
"There is no end to innovation, and it is a great honor to be a member of the Party and see our work being appreciated by the nation," he said.
In 2009, China was planning to build its first supercomputing center in Tianjin. "We had nothing to work with at the time. Our office was unfinished. It was just white walls and concrete floor, even the electrical wires were all hanging in the open," Meng said. "We had to begin preparation and planning for the center in our employee dorm."
In those trying times, Meng and his team worked as scientists and construction workers. They renovated the computer room, installed electrical wires and carried the computer cabinets, each weighting over 1 metric ton, to their appropriate locations.
They would also crawl under the cabinets to install cables for hours at a time in a room temperature of nearly 40 C.
After building the machine, Meng and his colleagues needed to ensure the new supercomputer could maintain stable performance at all times. They would calibrate the machine during the day, and at night they slept on the cardboard shipping boxes that came with the cabinets.
After seven months of hard work, Tianhe-1 was launched in October 2010, and was recognized as the fastest supercomputer in the world from November 2010 to June 2011.It was the first time that a Chinese supercomputer had made it to the top of the Top500 list.
But skepticism and doubts soon followed, with some experts claiming that Tianhe-1 was just a "giant gaming rig" and had no real industrial applications.
These comments infuriated Meng, who swore to improve the machine's application and software compatibility. "Only when the machine has useful applications can China cement its position in the global supercomputing sector," he said.
At the time, Tianhe-1 couldn't run a widely used research software program for developing new pharmaceuticals, so Meng spent over 20 sleep-deprived days optimizing hundreds of thousands of lines of code to solve the issue. This paved the foundation for the supercomputer's large-scale application in drug discovery and design.
"I always tell my teammates, every effort we put in will benefit our country in the end," Meng said. His team later developed the technology for processing high definition images used for geological investigation for oil and gas, eliminating China's need to spend millions of yuan to rent calculating capacity from supercomputers in the US and Europe.
Tianhe-1 is now used in a wide range of fields including aerospace technology, climate forecasting, nuclear fusion and tackling COVID-19 via drug screening and cloud diagnosis, Meng said. Since its inception, the instrument has led to over 4,000 breakthroughs in science and technology.
"It is truly an instrument of national significance that culminates generations of hard work by Chinese scientists and engineers," Meng said.
"As we push the scientific frontiers, we as members of the CPC need to be brave, innovative and contribute more to our country's self-sufficiency in science and technology."
Institute of Physics of CAS has developed a programmable 30-qubit superconducting-circuit based quantum computer this week [].
As of now the quantum computing groups in China are
1. USTC - Photonics and superconducting-circuit
2. Alibaba, well they've developed it last month, and they're the inventor & world leader in "fluxonium" qubit based superconducting-circuit. All the other teams are using transmon system. Alibaba now have the highest fidelity in superconducting-circuit based quantum computer []. They're also a leader in quantum algorithms.
3. CAS - Topological and superconducting-circuit
4. Origin Quantum (Hefei) - superconducting-circuit [], they are a full-stack company (Hardware+Software+Control systems)
5. TuringQ (Shanghai Jiaotong spinoff) - Photonics [], they are also a full-stack company. Moreover, they've also developed photonics-based QPU and photonic chip.
6. Huayi Quantum (Tsinghua spinoff), they're working on ion-trapped quantum computer.
6. QBoson (Beijing) - Full-stack photonic quantum computing []
7. SpinQ (Shenzhen) - Full-stack quantum computing []
8. Guokaike/QUDOOR (Beijing) - Cold-atom quantum computing []
9. QUTRONIX (Shanghai) - Quantum Control systems []
10. Baidu - Quantum algorithms.
11. Huawei?
I'll briefly mention other groups around the world.
In America,
Superconducting Circuits - Google, IBM, Rigetti
Photonics - PsiQuantum []
Xanadu [] well they're based in Toronto
Topological - Microsoft
Ion-trap - Honeywell
IonQ, University of Maryland spinoff
Cold atom - QuEra, MIT spinoff []
ColdQuanta, University of Colorado spinoff
Atom Computing
England - Superconducting Circuits
Photonics
Ion-trap
Algorithms
Control systems - Oxford Instruments
all basically Oxbridge spinoffs
France - Cold atom - PASQAL
Photonics - Quandela, CNRS spinoff
Netherlands - Superconducting Quantum Processor Units -
Photonics -
Algorithms -
Control systems -
all basically TU Delft spinoffs
Finland - Superconducting Quantum Computer -
Austria - Ion-trap, University of Vienna spinoff -
Switzerland - Algorithms -
Control systems - Zurich Instruments
Israel - Superconducting Circuits & control systems - Quantum Machines -
Photonics - Weizmann & Quantum Source -
Algorithms -
Interestingly Japan and Germany hasn't been able to deliver any quantum computer so far, and both are using quantum computing systems from elsewhere. However, Japan has a leading quantum software company, Qunasys [], a University of Tokyo spinoff. In Germany, Planqc [], a Max Planck spinoff, is trying to develop cold atom quantum computer.
As of now the quantum computing groups in China are
1. USTC - Photonics and superconducting-circuit
2. Alibaba, well they've developed it last month, and they're the inventor & world leader in "fluxonium" qubit based superconducting-circuit. All the other teams are using transmon system. Alibaba now have the highest fidelity in superconducting-circuit based quantum computer []. They're also a leader in quantum algorithms.
3. CAS - Topological and superconducting-circuit
4. Origin Quantum (Hefei) - superconducting-circuit [], they are a full-stack company (Hardware+Software+Control systems)
5. TuringQ (Shanghai Jiaotong spinoff) - Photonics [], they are also a full-stack company. Moreover, they've also developed photonics-based QPU and photonic chip.
6. Huayi Quantum (Tsinghua spinoff), they're working on ion-trapped quantum computer.
6. QBoson (Beijing) - Full-stack photonic quantum computing []
7. SpinQ (Shenzhen) - Full-stack quantum computing []
8. Guokaike/QUDOOR (Beijing) - Cold-atom quantum computing []
9. QUTRONIX (Shanghai) - Quantum Control systems []
10. Baidu - Quantum algorithms.
11. Huawei?
I'll briefly mention other groups around the world.
In America,
Superconducting Circuits - Google, IBM, Rigetti
Photonics - PsiQuantum []
Xanadu [] well they're based in Toronto
Topological - Microsoft
Ion-trap - Honeywell
IonQ, University of Maryland spinoff
Cold atom - QuEra, MIT spinoff []
ColdQuanta, University of Colorado spinoff
Atom Computing
England - Superconducting Circuits
Photonics
Ion-trap
Algorithms
Control systems - Oxford Instruments
all basically Oxbridge spinoffs
France - Cold atom - PASQAL
Photonics - Quandela, CNRS spinoff
Netherlands - Superconducting Quantum Processor Units -
Photonics -
Algorithms -
Control systems -
all basically TU Delft spinoffs
Finland - Superconducting Quantum Computer -
Austria - Ion-trap, University of Vienna spinoff -
Switzerland - Algorithms -
Control systems - Zurich Instruments
Israel - Superconducting Circuits & control systems - Quantum Machines -
Photonics - Weizmann & Quantum Source -
Algorithms -
Interestingly Japan and Germany hasn't been able to deliver any quantum computer so far, and both are using quantum computing systems from elsewhere. However, Japan has a leading quantum software company, Qunasys [], a University of Tokyo spinoff. In Germany, Planqc [], a Max Planck spinoff, is trying to develop cold atom quantum computer.
What do you think of D-Wave Systems of Canada?
One of the six satellites on board of the Lijian-1 rocket launched on July 27 is a nanosatellite for quantum communication experiments.
For the first time in the world, this tiny satellite will implement real-time quantum key distribution between a nanosatellite and miniaturized ground base stations. It will lay the foundation for practical and low-cost quantum encrypted communication networks between spacecrafts and ground base stations. In the future, global quantum communication networks can be built with constellation of such low-cost nanosatellites.
Comparing to the Mozi (Micius) satellite (QUESS, quantum experiments at space scale), while only one sixth in weight, this satellite has light source that runs at the frequency six times higher. The timeliness of encryption key generation by this nanosatellite is 2 to 3 orders of magnititude higher than the Mozi.
For the first time in the world, this tiny satellite will implement real-time quantum key distribution between a nanosatellite and miniaturized ground base stations. It will lay the foundation for practical and low-cost quantum encrypted communication networks between spacecrafts and ground base stations. In the future, global quantum communication networks can be built with constellation of such low-cost nanosatellites.
Comparing to the Mozi (Micius) satellite (QUESS, quantum experiments at space scale), while only one sixth in weight, this satellite has light source that runs at the frequency six times higher. The timeliness of encryption key generation by this nanosatellite is 2 to 3 orders of magnititude higher than the Mozi.
China’s anti-corruption watchdog investigates minister leading tech drive
Xiao Yaqing, the minister who has overseen China’s drive to and manufacturing, has been placed under investigation in what may be the first surprise of the upcoming .The probe was announced on Thursday by the Central Commission for Discipline Inspection (CCDI), the country’s . Xiao was placed under investigation for “suspected violations of the law and discipline”, according to the CCDI announcement.
Since 2020, Xiao has led the , which is at the centre of President Xi Jinping’s drive to reduce Beijing’s reliance on the developed world for technology. The ministry also oversees China’s vast internet infrastructure.
Take heed, tidalwave:
Amendment to rules surrounding posting content
Recent discussions across a number of threads have prompted the active moderating team to reflect and develop some rules and guidelines for the nature of the community's posting of content -- including articles, videos, images and social media, which may be of questionable quality, and lower the...
www.sinodefenceforum.com
The explanation is that this agency the MIIT is responsible for China's tech self-sufficiency drive. This is the agency that initiated Made in China 2025, which is heavily opposed by the U.S. So the fall of this agency right before the NPC is a big deal. Of course this pertains to China's scientific and technological development, and has major symbolic significance right after the U.S. passed the $280 billion bill to shore up its own science capabilities against China in particular.
In my view frankly, both the U.S. and Chinese governments are working towards the same goal: To make sure Chinese tech isn't too powerful, and that China lags behind further. There are other sources besides SCMP that have also posted this news, for instance, the Financial Times has an article. But frankly the Financial Times article does not have as much detail.
Well frankly your view ain't worth sh*t. The 280$ billion that you have contently promoted is over 5 years, not per year. And that includes baseline yearly funding so the extra funding is even less. The US spends 800$ billion on their military. But china is still catching up with less than a their of finding.
The agency MIIT is not "falling". For all your views on china failing, they have made 7nm chips and you couldn't cope.
Did you get bored that Russia is not losing the war yet? What happened to all those game changers for the Ukrainians?