Marin Ivezic
In 1996 a 26-year-old physics graduate from the University of Science and Technology of China arrived in Vienna to begin doctoral work with Anton Zeilinger. Five years later Pan Jianwei went home, to a country with no quantum information program worth the name. He is now executive vice president of USTC, and his name is attached to four of China’s flagship results: the Micius satellite, the Beijing–Shanghai quantum key distribution backbone, the Jiuzhang photonic machines, and the Zuchongzhi superconducting processors.
China spent the next two decades turning that single return into a system, and the system worked. It also ran into a limit that money can’t move. We keep coming back to this history when we teach quantum-readiness planning, for one reason: the same constraint binds a national program and a bank, and money can hire capability but can’t grow it.
One Return, and the Pattern It Set
Pan’s method was simple enough to copy, and China copied it for twenty years. Send your strongest candidates to the best laboratories in the world, then bring them back with money, laboratory space, and a title. The people who came back are the hǎiguī, sea turtles, a pun on the Mandarin phrase for returning from overseas. In quantum technology, a very small group of them built almost everything.
Lu Chaoyang took his doctorate at Cambridge and returned to USTC in 2011 to lead photonic quantum computing; the Jiuzhang machines are his group’s work. Chen Yu’ao trained at Heidelberg and the Max Planck Institute of Quantum Optics before returning the same year, and he engineered the fiber backbone between Beijing and Shanghai. Zhu Xiaobo spent five years at NTT’s laboratories in Japan and joined USTC in 2016 to build the Zuchongzhi superconducting line. Duan Luming spent 15 years at the University of Michigan, where he co-invented the DLCZ quantum repeater protocol that carries his initial, then resigned his chair in 2018 for Tsinghua. Guo Guangcan visited Toronto in the early 1980s, came home to introduce quantum optics into Chinese physics, founded the CAS Key Laboratory of Quantum Information in 2001, and later co-founded Origin Quantum.
Almost every Chinese quantum researcher’s intellectual lineage reaches Pan or Guo within two or three steps. Roughly a dozen people acquired world-class training abroad and then built institutions around themselves at home. That concentration is the program’s strength and its exposure at once.
The Recruitment Machine
Patriotism alone didn’t bring them back. China built the most elaborate scientific recruitment apparatus any state has run, and it is still expanding.
The Thousand Talents Program, launched in 2008, set the template. Its published terms offered senior recruits a signing bonus of 1 million RMB, research start-up funding in the millions, housing support, spouse employment help, and school places for children. Quantum physics was among the named priority fields. A junior tier, Young Thousand Talents, targeted researchers under 40. A 2023 study in Science found that its returnees went on to out-publish comparable peers who stayed abroad, and the gap was widest in last-author papers – the slot that shows who is running the laboratory.
American scrutiny changed the packaging more than the practice. The program was folded into a quieter successor from 2021. In October 2025 China opened the K visa, a category for early-career science and technology professionals that requires no employer sponsorship and sets no annual cap, with quantum research named as a target field. The same autumn, the United States announced a $100,000 fee on new H-1B petitions.
Underneath the national schemes sits a layered structure that Chinese academics call the hat ladder. Academician status in the Chinese Academy of Sciences sits at the top and controls access to funding, candidates, and laboratory space. Below it is a tier of named senior fellowships, and below that a junior tier that qualifies a researcher for the tier above. Universities bid against each other for hat-holders hard enough that the Ministry of Education has issued notices telling them to stop.
The Pipeline
Scale is the second advantage. Georgetown’s Center for Security and Emerging Technology projected China awarding around 77,000 STEM doctorates a year by 2025, roughly double the American figure, though only a small fraction of those graduates work anywhere near quantum information. The Ministry of Education has pushed universities to open quantum information science programs, and more than a dozen now run specialized curricula. USTC started a dedicated doctoral program in quantum science and technology earlier this decade.
USTC is the center of gravity. It ranked first in the world for quantum physics output in the 2025 Nature Index, a ranking of institutions by their share of papers in a set of highly selective journals. It hosts the Hefei National Laboratory, and the companies that grew out of its laboratories – QuantumCTek, Origin Quantum, CIQTEK – sit within a short drive of the campus.
Other poles have formed. Tsinghua’s Institute for Interdisciplinary Information Sciences, founded by the Turing Award winner Andrew Yao, hosts Duan’s trapped-ion group. Zhejiang University fabricates superconducting chips and absorbed Alibaba’s quantum laboratory when the company closed it in November 2023. The Beijing Academy of Quantum Information Sciences took in Baidu’s quantum laboratory on the same terms in January 2024.
Two of China’s flagship technology companies closed their quantum laboratories within three months of each other and moved their researchers to generative AI. State institutions absorbed the equipment and much of the staff, so the people stayed in the field, and the private commitment that funds long hardware programs in the United States is not what holds China’s program up.
What the Machine Bought
Judged against its own goals, the program delivered. China leads in quantum communications by a wide margin. Micius, launched in 2016, enabled intercontinental quantum-secured communication via satellite relay and demonstrated satellite-based entanglement distribution over roughly 1,200 km between ground stations in China, and the Beijing–Shanghai link is the longest quantum key distribution backbone in operation anywhere. Chinese photonic experiments have produced sampling results no classical machine has matched. Chinese superconducting processors are built at qubit counts comparable to the leading American chips.
That is a real national capability, assembled in under two decades from a standing start, and the recruitment machine paid for it.
What It Couldn’t Buy
Set the geopolitics aside for a moment. The gaps that remain sit in one family, and it is the family that governs how soon a cryptographically relevant quantum computer arrives.
A cryptographically relevant quantum computer, or CRQC, is one large and stable enough to break the public-key cryptography protecting today’s networks. Getting there depends far less on raw qubit counts than on quantum error correction, the technique of combining many error-prone physical qubits into a single logical qubit reliable enough to carry a long computation.
Error correction
The milestone the field watches is the below-threshold result, where adding more physical qubits makes the logical qubit better rather than worse. Google published one in December 2024 with its Willow chip. Chinese groups have since reported below-threshold results of their own, and the sensible posture is to wait for independent replication before treating any of them as settled. We apply that posture to every laboratory and every vendor, in every direction.
Error correction is where recruitment runs out of road. A signing bonus relocates an engineer who already exists. The number of people who have run a fault-tolerant experiment end to end is small in every country, and no national program has found a way to enlarge that group quickly. State money redistributes those people rather than producing more of them.
Software and tooling
Origin Quantum’s chief scientist, Guo Guoping, has said publicly that China lacks the engineers and technicians to turn research results into products. There is no Chinese Qiskit and no Chinese PennyLane, no domestic framework that developers outside the originating laboratory have adopted at scale. The two organizations best placed to build one walked away in 2023 and 2024.
Openness
The third gap follows from the first two. Error correction and quantum software both improve through fast, open, argumentative iteration across laboratories and across borders. China’s program runs mostly on domestic resources, and the quantum export controls introduced in September 2024, covering dilution refrigerators and cryogenic components, pushed it further inward. Some of that insularity was imposed and some was chosen. Either way it slows the work the program most needs to speed up.
American Restrictions and the Same Talent Pool
The United States has been running the mirror experiment. Proclamation 10043, signed in May 2020, suspended visas for Chinese graduate candidates tied to designated universities in fields including quantum computing. The Justice Department’s China Initiative ran from 2018 to 2022 and opened far more cases than it converted into convictions; the collapse of the prosecution against MIT’s Gang Chen in January 2022 became its emblem. A 2023 survey of Chinese-descent scientists in the United States, published in the Proceedings of the National Academy of Sciences (PNAS), found large majorities reporting that they did not feel safe as researchers, and a substantial share weighing whether to leave.
None of this expanded the world’s supply of error correction engineers. It relocated some of them. Every restriction that pushes a researcher out of an American laboratory and into a Chinese one subtracts from one side and adds to the other, and the global total doesn’t move. Some restrictions are still necessary. All of them deserve an honest accounting of the second-order effect. Fault-tolerant quantum computing is not a talent market anyone can raid into abundance, and both governments are bidding for the same short list of people.
What This Means for Your Organization
Almost nobody reading this will hire a quantum error correction researcher, and nobody needs to. The lesson transfers anyway, and it transfers through a smaller, much closer version of the same shortage: the people who can actually execute a post-quantum cryptography (PQC) migration.
Capability is a training line, not a procurement line. China had the money, the mandate, and two decades, and it still couldn’t buy past the engineering gap. An enterprise with a smaller budget and no mandate will meet the same thing at its own scale. Post-quantum readiness gets paid for in staff hours and skills, and the tooling arrives second.
The scarce roles sound ordinary. Nobody is short of people who can define a cryptographic inventory in the abstract. The shortage is in people who can produce one from a real estate of applications, read a hardware security module’s key hierarchy, argue with a vendor about ML-KEM (formerly CRYSTALS-Kyber) support in a product roadmap, and sequence a cutover that doesn’t break payments on a Monday morning. Those skills are learnable in months. They aren’t hireable on demand, and the price of them is rising.
Claim literacy is a control. The judgment we just applied to below-threshold results is the judgment your team needs for every vendor deck. Physical qubits are not logical qubits, an announced roadmap is not a demonstrated result, and “quantum-safe” on a datasheet means whatever the vendor decided it means. A team that reads the difference will spend its migration budget several years more accurately than one that can’t.
The bench you need in 2030 is already on your payroll. China’s advantage came from candidates it sent abroad 10 and 15 years earlier. That logic runs the same way inside your own organization. The person who will lead your cryptographic migration at the end of this decade is probably a PKI administrator, a network engineer, or an application security lead today, and the decision in front of us is whether we train them this year or advertise for them in 2029 alongside everybody else.
Where This Fits
Boards take the timeline signal from the China story. A state program of this scale, still short in error correction, is one input into how urgently the migration deadline should be read.
Practitioners take the workforce signal, and it is the one with a budget attached. Capability gets grown on a schedule, by people somebody decided to train.
Quantum Academy builds the second kind. Our certification programs are written for practitioners and for the executives who fund them, and they are organized around the work an organization has to do – cryptographic inventory, migration sequencing, vendor assessment, and program governance. The current programs and their entry requirements are listed at quantumacademy.com/.
For the migration methodology itself, pqcframework.org is the open reference. For the full analysis this article draws on, including the hardware programs and the investment picture, Marin Ivezic’s China’s Quantum Ambition series at PostQuantum.com goes considerably deeper.