# Quantum Is Accelerating

A mid-September 2026 field report on machines, megawatts and the map: where the United States and China stand, and why the record still matters more than the qubit count.

Canonical HTML: https://www.livetradingnews.com/quantum-is-accelerating
Last modified: 2026-09-16

---

By Shayne Heffernan · 2026-09-16
Tags: $IBM, $GOOGL, $IONQ, $RGTI, $QBTS, $QUBT, $NVDA, $ORCL, $MSFT, $INTC, $GFS, $BLK, $HON, quantum computing, post-quantum cryptography, QUOPS, harvest now decrypt later, ML-DSA, NIST, Armature L1
Signed: ML-DSA-65, anchored on Armature L1.
Nothing in this article is investment advice.

Quantum is accelerating. Not toward a machine that breaks RSA next quarter, which is still five orders of magnitude away on the first honest cross-platform yardstick the field has ever had, but toward foundries, clouds, logical qubits and government deadlines that are already fixed. Two clocks are running. Only one of them is slow, and it is not the one that decides what an institution has to do this year.

Quantum computing has a public-relations problem that looks, from a distance, like prudence. A new benchmark lands, the numbers are four or five orders of magnitude short of utility, a lab says fault tolerance is still years away, and the tape files the whole thing under science project. That reading is the noise. The work is not waiting.

## Two clocks, one noise

The first clock runs on a cryptographically relevant machine, a device that can factor the keys the world still uses to move money, sign treaties and keep hospital records closed. According to Sandia's QUOPS yardstick, published in the second week of September 2026, that machine is a long way off. Quantinuum's Helios-1 led the field at a Q score of 1,504. Google's Willow sat at 216. IBM's ibm_boston at 204. The same paper put factoring RSA-2048 near 250 million on that scale.

Five orders of magnitude is not a rounding error. It is also not a reason to leave the second clock alone.

The second clock is harvest-now-decrypt-later. Anything encrypted today under algorithms a future machine will break is already a letter to an adversary who can afford to wait. $BLK (NYSE: BLK) widened the quantum-risk language in its Bitcoin and Ethereum filings. NIST has dates: classical public-key cryptography deprecated by 2030, disallowed by 2035. The European Union wants high-risk systems migrated by 2030. The UK's NCSC wants priority migrations done by 2031 and the rest by 2035. Australia is aiming at 2030. The NSA's CNSA 2.0 clock is already on institutional calendars.

Those dates do not move because a Q score is short. That is the whole argument, and it is why the second clock, not the first, is the one a bank, a court or a ministry has to answer to in 2026.

![Two clocks: Sandia QUOPS scores for the leading quantum machines against the Q score needed to factor RSA-2048, and the fixed government migration deadlines that do not move when a benchmark is short](https://livetradingnews-media.nyc3.digitaloceanspaces.com/media/2026/09/16/cmpgg3-27fed64a7c8d69e2.svg)

*The machine is far. The deadlines are not. Sources: Sandia QUOPS, September 2026, and the published NIST, EU, NCSC, Australian and NSA timetables.*

> The hard problem was never intelligence. It is that banks, hospitals, courts and governments are being asked to let software act on their behalf with no way to establish who decided, on what basis, or whether the record will still read in a decade.

Substitute a quantum-relevant adversary for software and the sentence still holds. KXCO was built for that sentence. The Round Table is the room in which a human and a machine read the same sourced facts. [Armature L1](https://chain.kxco.ai) is the settlement layer that was post-quantum from its first block. [Ontology Live](https://kxco.ai/ontology-live) is the public map. The rest of this is the field report on the machines, the megawatts and the map.

## What actually shipped this season

Ignore the supremacy press releases and look at the engineering. Mid-2026 to mid-September 2026 produced more industrially useful quantum work than several prior years of conference talks.

### Cross-platform numbers, at last

QUOPS, the Quantum Universal Operations Performance System, was built at Sandia with Quantinuum and $NVDA (NASDAQ: NVDA) in the room. It is the first yardstick that puts trapped-ion and superconducting processors on one page. Q measures the size of a reliable circuit. Omega measures operations per second.

Helios-1, a 98-qubit trapped-ion machine, won on circuit size. Willow won on speed, posting on the order of 20 million operations per second against Helios-1's few hundred, per the same paper. IBM's superconducting entry sat near Willow on Q and well below it on throughput. A logical-qubit run on Helios scored 40. The paper's blunt conclusion is the one that belongs in a briefing: every system on the page is roughly a hundred thousand times short of the Q needed for useful scientific work.

That is the honest present tense. It is not a stop sign. A field that can finally be measured across platforms is a field where excuses have to get better.

### Machines and architectures

Nighthawk r2, from $IBM (NYSE: IBM), put 120 programmable qubits on a chip that spends most of its 458 on-chip qubits on coupling and fast reset. Tunable couplers drop relaxation from 200 microseconds to 25 nanoseconds on command. Throughput is reported at 100,000 circuits per second, per IBM, a twenty-five-fold gain over Heron. IBM also bolted two modular cryogenic units together in August and held the pair below 15 millikelvin, more than 180 times colder than deep space, in under five days.

That refrigerator is the unglamorous half of the 2029 Starling target: about 200 logical qubits running a hundred million gates. Switzerland will take the first IBM Quantum System Two on the continent. Yonsei in Seoul is pulling a 127-qubit Eagle and putting Nighthawk in its place in November, the only Nighthawk site outside IBM's own halls.

Willow, from $GOOGL (NASDAQ: GOOGL), 105 superconducting qubits, remains the error-correction reference from the December 2024 below-threshold result. Reinforcement-learning calibration through 2026 improved logical stability and pushed surface-code memory errors down. Google has also opened a second research line in neutral atoms. PsiQuantum, the photonic all-or-nothing bet, took the largest quantum-computing award of the summer at $125 million, per DARPA, as one of two firms in Stage C of the Quantum Benchmarking Initiative, and sits on a CHIPS letter of intent as well.

Quantinuum, majority owned by $HON (NASDAQ: HON), runs Helios, the trapped-ion machine an independent lab was willing to grade in Nature. Sandia confirmed the accuracy claims. Average two-qubit fidelity has been published north of 99.92 percent. Helix, the error-correction architecture, was validated on Helios in September: logical memory, logical Clifford work and logical entanglement across codes each beat the physical-level result without post-selection.

The company signed a $100 million CHIPS agreement for domestic ion-trap manufacturing with $GFS (NASDAQ: GFS) on 300-millimetre wafers, according to the Department of Commerce award list. Helios is slated for $ORCL (NYSE: ORCL) Cloud, and Singapore will be the first country outside the United States to host one.

$IONQ (NYSE: IONQ), having bought SkyWater as a foundry, launched Superion 256, trapped first ions in prototypes, and is talking about Superion 10K. It reported breakeven quantum error correction with qLDPC codes on a Tempo engineering system. Second-quarter revenue printed $80.1 million, per its own quarterly report.

Around those names the rest of the field kept moving. The 108-qubit Cepheus-1 from $RGTI (NASDAQ: RGTI) is generally available as twelve nine-qubit chiplets. Infleqtion guided 2026 revenue near $45 million, per company guidance, and is aiming at a 30-logical-qubit demonstration. Q-CTRL in Sydney ran a 100-qubit Fourier transform on IBM Heron, the largest on hardware to date. Oxford measured error correction across a photonic link between two ion processors two metres apart, the first correction to cross a network boundary. Finland's IQM put nine logical qubits on a delivery schedule for LUMI-Q. Nvidia shipped CUDA-Q Logical so that an algorithm, an error-correction scheme and a hardware design can be planned together before a machine is built, and a national lab cut one planning job from five months to three weeks.

That is industrialisation. It is not a pause.

![Five qubit modalities, the companies building each one, the physical input each is short of, and the wall-plug power bill they all converge on](https://livetradingnews-media.nyc3.digitaloceanspaces.com/media/2026/09/16/cmpgg3-d13de877203a1f67.svg)

*Five routes to a qubit, five different physical shortages, one shared electricity bill. Source: KXCO Round Table, 16 September 2026.*

## The United States: compute, capital and watts

The United States still fields the deepest commercial quantum sector on earth. The reason is not a single qubit record. It is the combination of three things China does not replicate easily: a public market that will list a trapped-ion company, a national laboratory system that will grade a vendor in Nature, and a foundry policy that now writes cheques measured in hundreds of millions.

According to the Department of Commerce, the 2026 CHIPS quantum round put just over two billion dollars against nine names and took minority equity. IBM took about a billion. GlobalFoundries took $375 million. $QBTS (NYSE: QBTS), Rigetti, PsiQuantum, Quantinuum, Atom Computing and Infleqtion took $100 million each. Diraq, the Australian silicon-spin house, took up to $38 million to put qubits on an American process.

DARPA's Quantum Benchmarking Initiative is the other filter. It is designed to kill architectures that cannot show a path to utility, which is why a Stage C award is worth more than a keynote.

On compute, the American stack is plural on purpose. Superconducting at IBM, Google and Rigetti. Trapped ions at Quantinuum and IonQ. Photonics at PsiQuantum. Neutral atoms at Infleqtion, QuEra and Atom Computing. Silicon spin arriving via HRL, the Malibu lab IBM agreed in July to buy from Boeing and GM, and via the Hitachi, $INTC (NASDAQ: INTC) and AIST line on Intel 18A. $MSFT (NASDAQ: MSFT) remains the software and topological wager. That plurality is a hedge. It is also a power bill.

American power is the binding constraint that quantum shares with AI, and it is the one the qubit essays skip. A superconducting stack lives at 10 to 35 millikelvin. Cooling is most of the wall-plug draw of a near-term machine. IBM's new modular fridges are the size of rooms. A future logical machine at a few thousand logical qubits has been sketched, depending on architecture, anywhere from under a megawatt to more than a hundred.

That is still small against a GB200 hall. It will not stay small if error correction forces the physical-to-logical ratio into the hundreds and the control stack scales with it. The interconnect queue that already throttles American AI campuses will throttle American quantum campuses the moment someone tries to put a utility-scale photonic plant and a hyperscale training cluster on the same substation. Chicago and Milpitas, Brisbane and the national labs are not abstract pins on a slide. They are bids for electrons.

What the United States leads is not raw announced funding, a column China dominates. It is the conversion of funding into graded machines, listed companies and a software layer the rest of the world actually uses. Qiskit, CUDA-Q, Azure Quantum and Braket are rails. Rails compound.

## China: a stack, not a copy

The American habit of describing Chinese quantum as a photocopy of American quantum is the same habit that misread Chinese AI. China is not trying only to hang a chandelier that looks like IBM's. It is trying to own the stack that turns a laboratory result into a factory, a cloud and a standard the domestic system can run without a licence.

On photonic compute, USTC's Jiuzhang line remains the national instrument. Jiuzhang 4.0, according to the Nature paper published in May 2026, used thousands of modes and more than three thousand photons. China Daily put the hardest sample at 25 microseconds against a classical estimate that, if you accept the comparison, leaves El Capitan in geological time.

In June, Tianyan-P2000, built by China Telecom Quantum Group with Jiuzhang in Jinan under CAS guidance, went onto the Tianyan cloud with a claimed 2,682-photon control, per the operator, and a 29-microsecond run against a 16-billion-year classical story.

Photonic machines run at room temperature. That is a power story as much as a physics story. Superconducting chandeliers drink cryogenics. Photon tables drink lasers and detectors. China is building both.

On superconducting compute, Origin Quantum in Hefei is the commercial flag. Origin Wukong, the 72-qubit third generation, has been on the cloud since January 2024, and per the operator has taken tens of millions of remote visits from more than 160 countries and completed hundreds of thousands of tasks. In May 2026 the fourth generation, Origin Wukong-180, went live on a single-core 180-qubit chip with domestically built measurement, environment and operating-system layers. Origin has booked a first overseas sale of Chinese quantum compute.

In early September the same Wukong hardware hosted coherent quantum routers for bucket-brigade QRAM: 98 percent transmission on a single router and 93 percent on a two-layer network, per results published in Physical Review X. Zuchongzhi 3.1, the 105-qubit academic cousin, prepared genuine cluster-state entanglement across 95 qubits. That is not a press department inventing a lead. It is a cluster that can put machines on a network and keep them there.

The industrial layer underneath is what Western coverage still underweights. Per contemporaneous local reporting, Origin raised on the order of 3 billion RMB in pre-IPO money earmarked for 10,000-qubit superconducting pilot lines. Turing Quantum and CHIPX in Wuxi run a photonic-chip pilot that cut some development cycles from months to weeks. Logical Qubit is building a Hangzhou cloud around logical-qubit milestones. Hyqubit out of Tsinghua has a 100-plus ion prototype at China Mobile. WeKe Quantum Light raised a 100-million-yuan pre-A on a glass ion-trap route after a Tsinghua group held 531 ions with coherence measured in hours.

China Mobile put high-dimensional GHZ and W states on a 16-millimetre silicon photonic chip in September. Neutral-atom startups have multiplied in Shanghai. Provincial quantum funds are now ordinary instruments, and Sichuan's 1 billion RMB book is one example. The 15th Five-Year Plan names fault-tolerant universal machines and scalable special-purpose machines as explicit goals.

On power, China is not fighting the same interconnect queue. It can add generating capacity at a pace no Western planning process matches, and it can put a photonic floor and a superconducting hall next to that capacity without a five-year interconnection study. That does not make the physics easier. It makes the industrialisation of the physics cheaper. Combined with a domestic post-quantum competition running on its own calendar, the result is a stack that does not need an American signature to operate.

Treat benchmarks as weather. Treat the factory, the cloud and the power plant as climate.

## Europe and the United Kingdom

Europe now leads the world on confirmed quantum funding even though it does not lead the world on listed hardware champions. According to Quantum Insider's Q3 2026 book, confirmed government money stood at $40.9 billion globally, of which Europe held $16.16 billion confirmed against Asia-Pacific's $15.92 billion. The announced column is a different planet: Asia-Pacific's $146.71 billion announced is mostly a Chinese figure, per the same tally. Europe's advantage is money that has actually been allocated, and a political machine that knows how to turn a flagship into a procurement.

The EU Quantum Flagship, EuroHPC systems, EuroQCI and the 2026 Quantum Computing and Simulation Roadmap are the official architecture. Member-state hardware is real: IQM in Finland, Pasqal in France, AQT in Austria, QuEra partnerships, Oxford Ionics and Quantinuum's Cambridge half in the UK, SAXON Q in Germany taking orders for a 128-qubit diamond box that plugs into a wall. The Commission's own count of machines by headquarters still shows the United States ahead in superconducting boxes and Europe dense across ion, cold-atom and photonic niches. Switzerland gets System Two. The Netherlands remains a components and optics power. France and Germany write industrial policy the way they write it for nuclear and rail.

The United Kingdom punches above its fiscal weight because it kept Cambridge Quantum in the merger that became Quantinuum, kept a serious ion and photonic bench, and wrote an NCSC migration timetable with dates a bank can put in a programme plan: discovery by 2028, priority upgrades by 2031, critical systems by 2035. That is not a machine. It is a forcing function. Forcing functions are how middle powers stay in a race they cannot win on capital expenditure alone.

## Asia beyond China

Japan called 2025 the first year of quantum industrialisation. Fujitsu and RIKEN have a 256-qubit superconducting line. Hitachi, Intel and AIST were selected to design silicon spin processors on Intel 18A, aiming at a 100-qubit error-corrected prototype by fiscal 2028 and a cloud testbed in 2027. IMS switched on Shunkai, a full-stack neutral-atom machine. METI is funding a quantum operating-system wager with KDDI and Jij, because Tokyo would rather write the interface than wait to buy it. Post-quantum migration is being planned toward 2035, with a detailed roadmap due in fiscal 2026. Japan's power system is tight. That matters more for AI halls than for the first logical machines. It will matter for both by the 2030s.

South Korea is replacing Eagle with Nighthawk at Yonsei, running its own KpqC algorithm set beside NIST, and targeting a national post-quantum transition by 2035 with pilots through 2028. The strategy reads less like build the chandelier and more like own the transmission. Repeaters and a regional quantum network are the distinctive Korean bets. That is a coherent middle-power choice.

India approved the National Quantum Mission at just over 6,000 crore rupees, per the government announcement, with hubs at IISc Bengaluru, IIT Madras, IIT Bombay and IIT Delhi. QpiAI has shown a 25-qubit machine. QNu Labs works the communications layer. C-DOT and Sterlite demonstrated a quantum-secured telecom link. MeitY and CERT-In published a migration white paper, and a 2026 task force pushed inventories and accelerated timelines for finance, defence and health. Rigetti booked an $8.4 million C-DAC order, per its own disclosure. The honest sentence is that India is late on hardware capital and early enough on policy that it will not be absent from the 2030s.

Singapore is doing what Singapore does: buy the best machine it can host and train people against it. Prime Minister Lawrence Wong's RIE 2030 plan named quantum a primary growth area. The country will be the first outside the United States to host Quantinuum's Helios. Anyon is building superconducting systems on the island. A workforce pact with Quantinuum and SIT is already running. For a city-state, sovereign access plus skills is a complete strategy.

Australia has spent a quarter-century on silicon spin and now has the commercial base to show for it. Diraq is on the CHIPS list. Silicon Quantum Computing sells a QPU. PsiQuantum's Brisbane facility is one of the two physical bets that company has placed on a million-qubit photonic machine. Q-CTRL is the control-layer export. Emergence Quantum and AirTrunk announced in September 2026 a partnership to put cryogenic cooling into hyperscale halls, the point at which Australian LNG and hydrogen know-how meets the millikelvin problem. Quantum at scale will require cryogenics at scale. That is not a slogan. It is an industrial observation from a country that already cools gas for a living.

## Middle East, Africa, Latin America

The rest of the map is not empty. It is thin, and the thinness is itself information.

In the Gulf the money is real and the theory of the case is sovereignty plus energy. Saudi Arabia, on the Vision 2030 clock, partnered with Pasqal for a 200-qubit machine aimed at energy, logistics and climate work. Aramco is not collecting posters. The UAE has hosted D-Wave's regional circuit and is writing a national encryption and quantum-secure programme. Qatar's Hamad Bin Khalifa University signed a three-year framework with $QUBT (NASDAQ: QUBT) for compute, sensing and communications. Israel is running Project Nexus, a tender for a domestically produced platform, because Jerusalem treats Q-day as a sovereignty event rather than a research milestone.

Iran, Turkey and Pakistan have laboratory activity that does not yet look like a national industrial programme. Confirmed Middle East and African government quantum funding is still in the low hundreds of millions against the tens of billions elsewhere, per the same Quantum Insider tally. Watts, in this region, are not the scarce input. Skills and supply chain are.

Africa's public programmes concentrate in South Africa, with university and national-lab work that has outlasted several fashion cycles. Egypt and a handful of other states have awareness programmes and scattered research groups. There is no African IBM and no African Origin. There is a generation of physicists who will either be hired by the stacks above or build a communications and sensing layer the continent actually needs. Sensing and quantum key distribution are the realistic near-term exports. A sovereign fault-tolerant machine is not.

Latin America has one national programme that looks like a programme: Brazil, with public quantum money now in the tens of millions of dollars and a community that has survived long enough to be named. Argentina's CNEA funded computing, materials and detection work in 2026. The rest of the region is researchers, not roadmaps. For a continent sitting on lithium, copper and hydropower, the inputs an AI and quantum build-out actually eats, that gap is strategic. Power without a stack is a raw-material story. Power with a stack is an industrial story.

Canada sits between the American system and a sovereign one. Waterloo remains a global research node. Xanadu listed as a photonic company in 2026. Anyon, Nord Quantique and Photonic took Quantum Champions cheques. Per the federal announcement, Ottawa put C$20.3 million into a thirteen-organisation defence quantum hub led by Calgary, and has been buying cryogenic amplifiers for government testbeds. Canada will not out-spend Washington or Beijing. It can remain the place that trains the people both of them hire.

## The Round Table, Ontology Live and the post-quantum record

Most of the industry still points a model at a white paper and asks it to guess when Q-day arrives. That produces sentences. It does not produce a migration.

KXCO inverted the order. Post-quantum cryptography, identity and an immutable record first. Ontology second, as the schema that says what a key, a permission, a counterparty and a settlement mean. The Round Table in the middle, because everything else reports to it.

The Round Table is the engine. The ontology is the record it keeps. Documents, payments, trading and identity are four fronts of one operation. A document is not executed until the record recognises the authority behind the signature. No value moves without a named identity and a live permission. People, institutions and AI agents sit on one cryptographic standard, issued, revocable and provable. On Meridian the rule is already operational: the AI drafts and checks, the human runs the deal. Every decision that matters carries a name and a time.

Armature L1 was post-quantum from genesis. There is no ECDSA basement to escape. Signatures are ML-DSA-65. Key encapsulation is ML-KEM-768. SLH-DSA is available. Signature checking runs on-chain at precompile 0x0b, signed by every validator and tested live with negative controls on 15 September 2026: a valid signature returns 1, while a tampered signature, a tampered message and a wrong key each return 0. Finality is about two seconds.

Per NIST's Algorithm Validation Test System, the library was graded in September 2026 at 2,130 cases across ML-KEM, ML-DSA and SLH-DSA, with zero failures, issued as demo certificate A11025. That is a demo certificate against NIST-generated vectors. It is not a FIPS 140-3 module validation, and we say so. The KXCO Quantum Index, reviewed in mid-August, still has the rest of a thirty-name custody and payments set clustered near zero. That is the attestation deficit in a table.

Ontology Live at [kxco.ai/ontology-live](https://kxco.ai/ontology-live) is the public demonstration, applied first to the AI sector: hundreds of entities, sourced claims, ranked findings, and an Analyst Outlook layer that keeps its own dates. Use it the way the working guide says. Findings first. Outlook second. Then the graph. Click a claim. Read the source. Note both dates. The same method belongs on the quantum file, across foundries, cryogenics, detector vendors, labs, ministries and power interconnects, because a qubit count without those edges is a press release. The institutional argument sits at [kxco.ai/ontology](https://kxco.ai/ontology).

## Power: the unglamorous constraint

AI and quantum are argued about as if they compete for the same headline. They compete for the same electrons. That is the sentence that should govern a national strategy.

A modern AI rack already pulls what a small building used to pull. Liquid cooling is no longer a preference. It is the condition for the latest accelerators. National interconnect queues in the United States are multi-year facts. Ireland has had to ration campuses. China adds generation the way other countries add task forces. The Gulf sits on fuel and is buying machines to sit next to it. Brazil and parts of Africa sit on hydro and have almost no stack to plug in.

Quantum's present-tense power draw is modest. A near-term chandelier is a laboratory object, and cooling is most of its bill. Photonic tables skip the millikelvin plant and pay in lasers and detectors. Room-temperature diamond boxes, if they work as advertised, pay in standard AC.

The future tense is different. Error correction multiplies physical qubits. Control electronics multiply with them. Projections for a few thousand logical qubits already span two orders of magnitude depending on architecture. Integrated quantum and classical halls in the 2040s have been sketched at the scale of today's data centres. Anyone who tells you quantum is low power is describing a prototype, not a plant.

This is why Australia's cryogenic data-centre work with AirTrunk belongs in a compute essay, why Saudi's Pasqal box belongs next to Aramco rather than next to a university poster, and why the American CHIPS awards that mention foundries and lasers are more serious than the awards that mention only qubits. Power is a chokepoint. The ontology treats it as one.

## Questions people ask

### Is quantum computing actually accelerating in 2026?

Yes, on engineering and deployment rather than on cryptographic threat. Between mid-2026 and mid-September 2026 the field gained its first cross-platform benchmark, error correction that beat the physical-level result without post-selection, modular cryogenic units held below 15 millikelvin, a CHIPS round of just over two billion dollars across nine companies per the Department of Commerce, and the first error correction measured across a network boundary. Utility-scale machines remain years away.

### When will a quantum computer break RSA-2048?

Not on any timeline the published benchmarks support. According to Sandia's QUOPS yardstick of September 2026, the leading machines scored 1,504 for Quantinuum's Helios-1, 216 for Google's Willow and 204 for IBM's ibm_boston, against roughly 250 million on the same scale for factoring RSA-2048. That is about five orders of magnitude. The migration deadlines are set by regulators, not by that gap.

### What is harvest-now-decrypt-later?

It is the practice of capturing encrypted traffic today and storing it until a machine exists that can break the encryption. It makes the threat present-tense even though the machine is not. Anything with a long confidentiality life, such as medical records, legal files, state secrets and settlement records, is exposed the day it is transmitted, not the day the machine arrives.

### Which country leads quantum computing?

It depends which column you read. The United States leads the conversion of funding into graded machines, listed companies and the software rails others use. China leads announced funding and the industrial stack, with photonic and superconducting services on one domestic cloud and power it can add quickly. Europe leads confirmed, allocated government money at $16.16 billion against Asia-Pacific's $15.92 billion, according to Quantum Insider's Q3 2026 tally.

### What are the post-quantum migration deadlines?

NIST deprecates classical public-key cryptography by 2030 and disallows it by 2035. The European Union wants high-risk systems migrated by 2030. The UK's NCSC wants discovery by 2028, priority upgrades by 2031 and critical systems by 2035. Australia is aiming at 2030. The NSA's CNSA 2.0 timetable is already on institutional calendars.

## The close

Mid-September 2026 is a loud room. QUOPS says the machines are short. Origin and Jiuzhang say China can put photonic and superconducting advantage-class services on one cloud. IBM says the refrigerator now comes in modules. IonQ says the foundry can tape out a 256-ion chip. DARPA says only two bets have earned Stage C. Ministries say 2030 and 2035 as if dates were physics. The useful act is to separate the noise from the work.

The work is this. Logical qubits are arriving in dozens, not thousands. Cross-platform benchmarks now exist, which means excuses will have to get better. The United States leads the commercial conversion of machines into products and rails. China leads the conversion of results into a domestic stack with power attached. Europe leads money that has actually been allocated and a regulator who will force a migration. The UK, Japan, Korea, Singapore, Australia, Canada and Israel are running coherent middle-power strategies. India is building late and in public. The Gulf is buying. Africa and most of Latin America are watching a race that will set their cryptographic future whether they join it or not.

None of that is a reason to wait on the second clock. A record that will not verify after the classical primitives die is a record an institution cannot take to court.

KXCO built the rails in a particular order on purpose. Security first. Settlement first. Then meaning. Then the room in which a human and a machine sit over the same facts and the human still decides. The Round Table is that room. Ontology Live is the public proof. The library NIST graded is the proof that the signatures were not a slide.

Ignore the noise. Quantum is accelerating. Accountability is what has failed to speed up. That is the entire opening. The rest is implementation.

Stocks mentioned in this article: $IBM, $GOOGL, $IONQ, $RGTI, $QBTS, $QUBT, $NVDA, $ORCL, $MSFT, $INTC, $GFS, $BLK and $HON.

Further reading on Live Trading News: the [KXCO centre](https://www.livetradingnews.com/center/kxco), the [AI Stocks centre](https://www.livetradingnews.com/center/ai-stocks) and the [stocks desk](https://www.livetradingnews.com/center/stocks). The author archive is at [livetradingnews.com/author/shayne-heffernan-phd](https://www.livetradingnews.com/author/shayne-heffernan-phd).

Shayne Heffernan, Ph.D., is the founder of Live Trading News, the KnightsBridge Group, Knightsbridge Law and the KXCO.ai ecosystem spanning post-quantum cryptography, identity, attestation and enterprise ontology. His signed long form is published at [shayneheffernan.com](https://www.shayneheffernan.com).

*This briefing is general information and analysis. It is not personal investment advice and it is not an offer of KXCO securities. Hardware scores cite the publishing labs or the Sandia QUOPS comparison as of mid-September 2026. Government funding totals cite contemporaneous tallies and will move.*

---

This Markdown mirrors https://www.livetradingnews.com/quantum-is-accelerating. The HTML page is canonical.
Site index for AI clients: https://www.livetradingnews.com/llms.txt