China's supercomputer breakthrough uses 37 million processor cores to model complex quantum chemistry at molecular scale — Sunway fuses AI and quantum science
A new breakthrough for China?
Normally, simulations for scientific research are performed on supercomputers as they require tremendous compute throughput. There are also types of research — such as simulation of quantum behavior of molecules with exponentially more interacting states — that require quantum computers to simulate them, or simplifications to make the task suitable for modern supercomputers. However, Chinese scientists from Sunway have successfully used an AI model and an existing Oceanlite supercomputer to model complex quantum chemistry at the scale of real molecules, which is both a scientific and technological breakthrough, reports VastData.
A quantum state in quantum mechanics — described by a wavefunction (Ψ) — determines all possible configurations of a quantum system, such as the positions, spins, or energy levels of particles like electrons in a molecule, along with their probabilities. Modeling it is challenging because the state space grows exponentially with the number of particles, making it impossible (and not feasible) to simulate on classic supercomputers that we use today. To that end, scientists use a variety of approximation methods to simplify the quantum equations while preserving accuracy to describe molecular structures, reactions, and energies. However, the scaling of existing methods that approximate the wavefunction is limited to small molecules.
To study many-body quantum systems with strong electron correlations (e.g., tens of electrons, over 100 spin orbitals, etc.), several years ago, physicists proposed using modern machine-learning surrogates, such as neural-network quantum states (NNQS), to approximate all the possible configurations and motions of electrons within a molecule. This method promises to wed AI scalability with quantum accuracy for research that is currently impossible using traditional methods.
To conduct their experiment with 120 spin orbitals, researchers developed their own NNQS framework. Their simulation trained a neural network to approximate the molecule's wavefunction, determining where electrons are most likely to be. For each sampled electron arrangement, the system calculated the local energy and adjusted the network until its predictions matched the molecule's true quantum energy pattern.
The proprietary NNQS that was tailored for China's Oceanlite supercomputer, based on a 384-core Sunway SW26010-Pro CPU, which supports FP16, FP32, and FP64 data formats and features a very unique architecture tailored for HPC rather than for AI. In particular, they had to keep in mind how the SW26010-Pro parallelizes workloads and handles data.
The researchers have designed a hierarchical communication model where management cores handled coordination between processors and nodes, while millions of 'lightweight' 2-wide compute processing elements (CPEs) featuring a 512-bit vector engine performed local quantum calculations. In addition, they created a dynamic load-balancing algorithm to ensure that uneven computational loads did not leave any cores idle.
The scientists ran their code on 37 million CPE cores and achieved 92% strong scaling and 98% weak scaling, a high level of efficiency for such a scale, which highlights that the developers have found a near-perfect synchronization between software and hardware, a major accomplishment for China's supercomputer community. Also, to date, the simulation of molecular systems with 120 spin orbitals is the largest AI-driven quantum chemistry calculation ever performed on a classical supercomputer, marking a breakthrough for PRC's AI and quantum industries.
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Without any doubt, the achievement demonstrated that NNQS can be used for quantum physics research on modern supercomputers. However, it is impossible to find out whether it is efficient to use an exascale supercomputer like Oceanlite for AI-based quantum physics research, both in terms of effort and power.
Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.
Anton Shilov is a contributing writer at Tom’s Hardware. Over the past couple of decades, he has covered everything from CPUs and GPUs to supercomputers and from modern process technologies and latest fab tools to high-tech industry trends.
-
twin_savage I'm seeing some software vendors and "experts" push DNN surrogate models to solve traditional FEA problems and it is a mess; they are very often inaccurate and don't produce an error function like a PCE surrogate would so they don't even quantify how inaccurate the results they spit out are.Reply
I can't imagine the surrogate model for spin orbitals is any better. -
ejolson After constructing quantitative neural networks at a much smaller scale the observation here is they are able to create good approximations from less data but as the quality of data increases the quality of the approximations do not get appreciably better. It could just be me.Reply
At any rate, Oceanlite turned out to be a good research platform for people doing interesting work. It also demonstrates just a sampling of what DEC might have created if not destroyed years ago by Compaq for its sales team.
I'm disappointed Sun was destroyed in a similar way, but that's a different story. -
Constellar Look you guys, I understand a lot about the universe we live in. Heck, I even understand that the millibytes of the conjoined quantum wells were met by the friend of the second-order integral that was known for the shattering of the twin seas over the seven eviscerating horizons, but I never, ever, read something as incomprehensible as this article...Reply
Congrats if you were able to keep up with half of it lmao... -
reaperofchange China has over a billion people, and has been tought everything we the usa has educated them with that they use now. Thanks to that education, our research, and efforts that provided china with knowledge necesary to even begin to write a research paper that sounds convincing enough, all these claims made everyday of breakthroughs wouldnt be possible if it wasnt for the plethora of wisdom that china was blessed with that leaked out of the usa and other country's. Now we just sit and wait for the zero recognition that will arrogantly not be announced because the ego's are to large to consider even acknowledging that the amount china has personally done, and used up in resources just to gain the data in legitimate instances compared to what America had to that is now in the hands of china despite what they will admit, is not even close. The efforts done in America by researchers from all over the world, working to bring the very tech everyone takes for granted today and years to come, will outway that of china for a long time. Its great china is finally stepping in to do some of the heavy lifting if this all turns out to actually be that rather than just claim after claim. We are a third the size of china yet it keeps bragging it does this or that better than America, but it should be pretty obvious that they should have been for much longer ago but has just reached that point to start. That shows its still got its work cut out for it and probably should start showing a bit more humbleness rather than always picking on the smaller kids just to feel big. Its been big but why has it been taking so long to help advance things? Maybe we in america can sit back and relax for a bit now, let china do some of the chores for a bit.Reply -
urn66 Reply
I have been repeating this for more than a decade as the dominos fell. Some people see a roadblock, other a challenge, still others a prompt to rethink the problem. China's supercomputing evolution ( and now quantum computing evolution) has followed the latter 2 paths. If you research several years of Top 500 Computer rankings, you ca see the ebb and flow, but more Top 5 and Top 1 rankings than most people would imagine.phead128 said:US export controls really jolted China to go all-in didn't they.
See truth from facts. -
urn66 Reply
American universities (and companies) recruit Chinese students/researchers/engineer because they are bright , well-educated, hard-working and creative. In fact, K-12 education in China is top rank and definitely above the USA, where academic rankings are actually falling.reaperofchange said:China has over a billion people, and has been tought everything we the usa has educated them with that they use now. Thanks to that education, our research, and efforts that provided china with knowledge necesary to even begin to write a research paper that sounds convincing enough, all these claims made everyday of breakthroughs wouldnt be possible if it wasnt for the plethora of wisdom that china was blessed with that leaked out of the usa and other country's. Now we just sit and wait for the zero recognition that will arrogantly not be announced because the ego's are to large to consider even acknowledging that the amount china has personally done, and used up in resources just to gain the data in legitimate instances compared to what America had to that is now in the hands of china despite what they will admit, is not even close. The efforts done in America by researchers from all over the world, working to bring the very tech everyone takes for granted today and years to come, will outway that of china for a long time. Its great china is finally stepping in to do some of the heavy lifting if this all turns out to actually be that rather than just claim after claim. We are a third the size of china yet it keeps bragging it does this or that better than America, but it should be pretty obvious that they should have been for much longer ago but has just reached that point to start. That shows its still got its work cut out for it and probably should start showing a bit more humbleness rather than always picking on the smaller kids just to feel big. Its been big but why has it been taking so long to help advance things? Maybe we in america can sit back and relax for a bit now, let china do some of the chores for a bit.
Suggesting China can only achieve because some magic knowledge is "gifted" to them or "stolen" by them from the USA is beyond arrogant and culturally biased; it is, dare I say, r*c*st.
Maybe you should get some facts. Maybe you should check the rankings of top universities in the world, top cited research papers, and see how the US, China and other nations rank. You are in for a surprise.
Good ideas can come from anywhere, and arrogance limits your own ability to think.
News Flash: China has been a civilized society for 5,000+ years and was ahead of the West for most of that time. The decline of our Imperial System and a period of Western and Japanese Colonialism slowed us down, but that is over. -
urn66 Sunway has a very creative team of scientists, engineers and coders, and Ince it is a shared computing resource center, it gets a diverse range of problems to solve. This demonstrates the merit of such institutions, whether they are university based, of like Sunway, a branch of a national computing center.Reply
OK, now lets see what others can do with the or a similar approach, science cannot wait another 5 years to integrate AI into large, complex modeling. -
reaperofchange Reply
It’s hard to ignore how political deals often serve hidden agendas, moving people and resources around for lobbying interests rather than genuine progress. This isn’t about labeling anyone as ignorant, it’s about acknowledging that the U.S., like many nations, has grown complacent. When people grow too comfortable with the illusion of safety and success, focus fades.urn66 said:American universities (and companies) recruit Chinese students/researchers/engineer because they are bright , well-educated, hard-working and creative. In fact, K-12 education in China is top rank and definitely above the USA, where academic rankings are actually falling.
Suggesting China can only achieve because some magic knowledge is "gifted" to them or "stolen" by them from the USA is beyond arrogant and culturally biased; it is, dare I say, r*c*st.
Maybe you should get some facts. Maybe you should check the rankings of top universities in the world, top cited research papers, and see how the US, China and other nations rank. You are in for a surprise.
Good ideas can come from anywhere, and arrogance limits your own ability to think.
News Flash: China has been a civilized society for 5,000+ years and was ahead of the West for most of that time. The decline of our Imperial System and a period of Western and Japanese Colonialism slowed us down, but that is over.
Despite its flaws, America once brought people together from across the world, offering opportunity and connection on a global scale. China, meanwhile, continues to paint the U.S. and the West as “warmongers,” yet its own record tells a harsher truth. Historically, China has been responsible for immense internal suffering, often through self-inflicted cultural and political disasters. These are moral concerns, not racial ones.
Whenever discussions about China arise, race is often injected into the debate, usually by China itself. The irony is that while it calls out racism abroad, it enforces one of the world’s most rigid systems of racial and cultural conformity. If China wants to be seen as a global leader, it must lead by example: through cooperation and shared progress, not propaganda and control.
America, for all its imperfections, remains a melting pot striving to overcome division. Much of the racial tension discussed today stems more from political opportunism than reality. True progress means moving beyond weaponized narratives and focusing on what unites, not divides.
Historically, even if every U.S. war were combined, the total human cost wouldn’t come close to China’s own domestic tragedies. That’s not an insult, it’s a sobering fact. And as history shows, arrogance often comes before decline. China has fallen behind before, and unless it learns humility, it will again.
The world doesn’t need another empire fighting for dominance. It needs nations willing to cooperate for stability and human progress. The average American doesn’t want war, they want peace, communication, and growth. It’s political ambition, not public will, that drives hostility.
If China truly wants global respect, it must abandon the pursuit of control and embrace genuine collaboration. Strength isn’t found in domination, it’s found in helping humanity move forward together. -
reaperofchange With that said, i commend china on the successes that may be constructive for technology and practices that progress it. I just hope those that are being made are for good reasons, not just for the purpose of deceptive, malicious, or belittling purposes.Reply
Most posts coming out of there in recent times are usually paired together with how it makes tech from others like they were made by "insert belittlement statement here" from such and such.
Why not just put it as "a new step to drive humanity forward" instead rather than condemning self to resorting to low blow tactics making one no better than the competition it is competing with.