Quantum Computing Explained Without the Math
Quantum computers use qubits, superposition, entanglement and interference to tackle a narrow set of hard problems. They are real but error-prone, and are not expected to replace ordinary computers.
// quantum computing
Quantum Computing Milestones Tracker
A dated, sourced log of quantum computing milestones since January 2023: processor and error-correction results, published company and government roadmap targets, and post-quantum cryptography standards and deadlines.
Researched and fact-checked by AI, with no human review. How we verify
25 entries from 29 sources. Checked for news weekly; last checked
What this tracker follows
This tracker logs dated milestones in quantum computing from January 2023 to October 2026. It follows three threads. The first is hardware and error-correction results from IBM, Google, Quantinuum, IonQ, PsiQuantum, Microsoft and university-led groups. The second is roadmap targets published by those companies and by governments. The third is the standards and deadlines for post-quantum cryptography, meaning cryptographic algorithms designed to resist attack by a quantum computer.
An event gets an entry if a source confirms what happened and when. The source must be a peer-reviewed paper, an official announcement or a report in an established trade publication. Much of the list concerns logical qubits. These spread information across many physical qubits so that errors can be caught. Performance figures and claims of quantum advantage are attributed to whoever made them. Quantum advantage means a quantum machine doing something classical computers cannot practically match. Disputes are noted, as with the critique published in Nature of Microsoft's topological qubit evidence and Microsoft's reply. Future targets are marked as scheduled and say who set them and when. NIST's 2035 date comes from a report that was still an initial public draft when checked on October 6, 2026.
Left out on purpose: funding rounds, stock listings, acquisitions, partnerships and bare qubit-count announcements. Also left out are theoretical resource estimates and outside forecasts of when quantum computers will become commercially useful. Vendors use different hardware and report different metrics, so the list is not a ranking.
Scheduled next
Targets set by the organizations named. Dates like these often move.
Scheduled
NIST's proposed date to disallow RSA and elliptic-curve public-key algorithms
NIST released a draft report, IR 8547, in November 2024. It proposes that quantum-vulnerable public-key algorithms such as RSA, ECDSA and Diffie-Hellman be disallowed in NIST's standards after 2035. It proposes that 112-bit-strength versions be deprecated after 2030. The dates are draft proposals, not final rules.
Scheduled
UK cyber agency's target for completing post-quantum migration
The UK National Cyber Security Centre set this target in guidance published March 20, 2025. It asks organizations to finish discovery and an initial plan by 2028 and carry out their highest-priority migrations by 2031. It asks them to complete migration of all systems, services and products by 2035.
Source: Timelines for migration to post-quantum cryptography, UK National Cyber Security Centre
Scheduled
U.S. federal deadline for post-quantum key establishment
Executive Order 14412 was signed June 22, 2026. It directs the Office of Management and Budget to issue guidance. The guidance is to require each agency to move high value assets and high impact systems to post-quantum cryptography for key establishment by December 31, 2030. It is to require the same for digital signatures by December 31, 2031.
Source: Securing the Nation Against Advanced Cryptographic Attacks, The White House
Scheduled
IBM's target for Starling, a fault-tolerant quantum computer
IBM said on June 10, 2025 that it plans to deliver Starling in 2029 at a data center in Poughkeepsie, New York. It said Starling would have 200 logical qubits able to run 100 million quantum operations. It listed the Loon (2025), Kookaburra (2026) and Cockatoo (2027) processors as steps. This is a company target.
Scheduled
Google's target for its post-quantum cryptography migration
Google security staff Heather Adkins and Sophie Schmieg set 2029 as the company's timeline for migrating to post-quantum cryptography. They set it in a March 25, 2026 blog post, citing progress in quantum hardware, error correction and factoring resource estimates. This is a company target.
Source: Quantum frontiers may be closer than they appear, Google
Scheduled
U.S. Energy Department's target for fault-tolerant quantum systems
The Department of Energy said on June 23, 2026 that its Quantum Genesis initiative includes a competition. It said the competition is to demonstrate, in 2028, fault-tolerant quantum systems with logical qubit counts in the low hundreds. It tied the initiative to a presidential executive order on quantum technology.
Timeline
IonQ opens orders for its 256-qubit Superion platform
IonQ announced Superion, its sixth-generation trapped-ion platform, and said the 256-qubit Superion 256 could be ordered, with customer deliveries in 2027. It said first ions had been trapped in prototype systems and that it expects fault tolerance in a lab setting in 2027. These are company claims.
IBM and partners release three quantum advantage claims
IBM and collaborators at the University of Chicago, Qedma and Algorithmiq released three papers. The papers claim calculations on IBM's 156-qubit Heron processor that classical computers cannot match. IEEE Spectrum reported none had been peer reviewed. It quoted an outside researcher urging caution on two.
Source: IBM Claims Quantum Advantage With New Validation Techniques, IEEE Spectrum
Nature publishes a critique of Microsoft's topological qubit evidence
The critique is a Matters Arising article by physicist Henry Legg of the University of St Andrews. Legg argued that transport data behind Microsoft's February 2025 Nature paper show its parity readout occurred in disordered regions that appear gapless. Microsoft Quantum disputed this in a reply published the same day.
Source: On the robustness of topological gap detection via transport, Nature
PsiQuantum starts construction at its Queensland site
PsiQuantum said it had broken ground and started construction at Moreton Bay Central in Queensland, Australia. It plans to build there what it calls the first utility-scale, fault-tolerant quantum computer. It said a cryoplant ordered in late 2024 is due for delivery in the second half of 2027.
Microsoft introduces its Majorana 2 chip at Build
Microsoft introduced Majorana 2 at its Build conference on June 2-3, 2026. It said the chip's qubits were 1,000 times more reliable than the previous generation, with a mean lifetime of 20 seconds. It set 2029 as its goal for a commercially relevant, scalable quantum computer. These are company claims.
Source: Introducing Majorana 2, Microsoft's next-gen quantum chip (Microsoft Build Live), Microsoft
IBM announces its Nighthawk and Loon processors
IBM announced Nighthawk, a 120-qubit processor with 218 tunable couplers. It expected to deliver Nighthawk to users by the end of 2025. It also announced Loon, an experimental chip it said shows the processor components fault-tolerant computing needs. It reported real-time error decoding in under 480 nanoseconds.
Quantinuum launches its 98-qubit Helios computer
Quantinuum announced the commercial launch of Helios, a trapped-ion machine with 98 physical qubits and a reported two-qubit gate fidelity of 99.921%. The company said it had demonstrated 48 error-corrected logical qubits on it. It said access is through its cloud service or an on-premises system.
Google reports its Quantum Echoes result and claims verifiable advantage
Google said its Quantum Echoes algorithm was 13,000 times faster than the best classical method on a leading supercomputer. It said the algorithm ran on the Willow chip and was described in a Nature paper. Google called the result the first verifiable quantum advantage. The advantage claim is Google's own.
NIST selects HQC as a backup post-quantum encryption algorithm
NIST chose HQC as a backup to ML-KEM. HQC is built on error-correcting codes, while ML-KEM relies on structured lattices. NIST called HQC the fifth algorithm selected in its post-quantum project. It said then that it expected a draft standard in about a year and a final standard in 2027.
PsiQuantum publishes its photonic chip platform in Nature
A peer-reviewed paper from PsiQuantum described photonic chips made in a commercial semiconductor foundry on a 300-millimeter process. It reported a two-qubit fusion fidelity of 99.22% and a chip-to-chip qubit interconnect fidelity of 99.72%. These are component benchmarks, not a working computer.
Source: A manufacturable platform for photonic quantum computing, Nature
Microsoft unveils Majorana 1 and claims topological qubits
Microsoft announced Majorana 1, a chip it said holds eight topological qubits and is designed to scale to a million. It called a Nature paper peer-reviewed confirmation, though that paper's abstract weighs both topological and trivial explanations. The claim is disputed; see the June 24, 2026 entry.
Source: Microsoft's Majorana 1 chip carves new path for quantum computing, Microsoft
Google reports error correction below threshold on its Willow chip
A Nature paper from Google Quantum AI reported on logical error rates. It said they fell by a factor of 2.14 each time the surface code distance rose by two. It reported this up to a distance-7 code using 101 qubits of a 105-qubit Willow processor. It said that memory outlasted its best physical qubit by a factor of 2.4.
Source: Quantum error correction below the surface code threshold, Nature
NIST finalizes its first three post-quantum cryptography standards
NIST released FIPS 203 (ML-KEM, from CRYSTALS-Kyber) for general encryption. It also released FIPS 204 (ML-DSA, from CRYSTALS-Dilithium) and FIPS 205 (SLH-DSA, from Sphincs+) for digital signatures. It said then that a draft of a fourth standard, FIPS 206 based on FALCON, was planned for late 2024.
Quantinuum expands its H2 computer to 56 qubits
Quantinuum said its H2-1 trapped-ion processor had been upgraded to 56 fully connected qubits. It said a random circuit sampling test was run with JPMorgan Chase, Caltech and Argonne National Laboratory. It said the test scored about 0.35 on a fidelity benchmark, over 100 times the score of Google's 2019 experiment.
Microsoft and Quantinuum report four low-error logical qubits
Microsoft said the companies ran its qubit-virtualization system on Quantinuum's ion-trap hardware. It said that in doing so they created four logical qubits with an error rate 800 times lower than that of the physical qubits. It said they ran more than 14,000 experiments without an error. The result is company-reported.
Harvard-led team runs circuits on up to 48 logical qubits
A Nature paper from a Harvard-led group described a processor built from up to 280 neutral-atom physical qubits. It said the processor encoded as many as 48 logical qubits and ran sampling circuits with 228 logical two-qubit gates. It reported that logical encoding with error detection improved algorithm performance.
Source: Logical quantum processor based on reconfigurable atom arrays, Nature
IBM debuts its 133-qubit Heron processor and Quantum System Two
IBM introduced Heron, a 133-qubit processor, at its Quantum Summit in New York. It said Heron cut errors up to five-fold compared with its Eagle chip. It said the first Quantum System Two, in Yorktown Heights, New York, had begun operating with three Heron processors. It also extended its roadmap to 2033.
IBM and UC Berkeley publish evidence of utility before fault tolerance
A Nature paper by IBM, UC Berkeley and others reported accurate results from an error-mitigated 127-qubit processor. It reported those results on circuits of up to 2,880 two-qubit gates where tested tensor-network methods failed. Flatiron Institute researchers later published a classical simulation they called more accurate.
Source: Evidence for the utility of quantum computing before fault tolerance, Nature
Google reports a larger error-correcting code outperforming a smaller one
A Nature paper from Google Quantum AI reported that a distance-5 surface code on 49 qubits had a logical error per cycle of 2.914%. It reported an average of 3.028% for distance-3 codes on 17 qubits. Both figures are from a 72-qubit processor. The gain was small, but the larger code had the lower error rate.
Source: Suppressing quantum errors by scaling a surface code logical qubit, Nature
Additional sources
- NIST IR 8547 (Initial Public Draft): Transition to Post-Quantum Cryptography Standards, National Institute of Standards and Technology, Computer Security Resource Center
- Reply to: On the robustness of topological gap detection via transport, Nature
- Interferometric single-shot parity measurement in InAs–Al hybrid devices, Nature
- Efficient Tensor Network Simulation of IBM's Eagle Kicked Ising Experiment, PRX Quantum (American Physical Society)
Update history
- Rewritten in shorter, plainer sentences. No facts were changed.
Articles on Quantum Computing
When Will Quantum Computers Break Encryption?
No one knows. Estimates run from a few years to a few decades, and a survey published in March 2026 put the average odds within 10 years at 28% to 49%. Migration deadlines run from 2029 to 2035.