Blockchain Quantum Fracture: What Will Stand the Test and What Remains a Forecast for Bitcoin

01.11.2026

A Comparative Analysis of the Vulnerability of Bitcoin, Ethereum, Litecoin, Monero, and ZCash to Shor and Grover Attacks—and a Framework of Eight Reviewable Studies Based on: JJ Kearney, CA Perez-Delgado, “Vulnerability of Blockchain Technologies to Quantum Attacks”.

Content

⚙️ Detailed research option

  1. A scene that frightens and deceives
  2. Theme, central question and four hypotheses
  3. Method: How to Distinguish Evidence from Prediction
  4. Mechanism: Shor, Grover, and the Special Fragility of Blockchain
  5. Resource Model: What 42 Minutes Really Mean
  6. Five Cases: From Bitcoin to ZCash
  7. Comparison and rating – with reservations
  8. Critical analysis: where the article is strong and where it is conditional
  9. A program of eight verifiable studies
  10. Practice: what to change and in what order
  11. Open questions
  12. Checklist before publication
  13. Sources
  14. Article working card

Annotation

⚙️ Detailed research option

This article answers a question that often gets lost in sensational headlines:  which specific elements of blockchain systems become insecure with the advent of a quantum computer of sufficient size?  Using a comparative analysis by Kearney and Perez-Delgado, we reconstruct the threat model for Bitcoin, Ethereum, Litecoin, Monero, and ZCash, categorize the assertions into proven, estimated, and predictive, and propose a program of eight verifiable studies.

The main conclusion: the bottleneck isn’t hashing or consensus per se, but  elliptic curve digital signatures  and  the public key availability window . Privacy changes the attack economics, but not the cryptographic strength. A unique catastrophic vector is the global public zk-SNARK parameter in ZCash, which allows for emission “out of thin air,” invisible to the network.

Keywords:  quantum cryptanalysis, Shor’s algorithm, Grover’s algorithm, blockchain, ECDSA, EdDSA, post-quantum cryptography, zk-SNARK, threat model.

1. A scene that frightens and deceives

⚙️ Detailed research option

The year is 2035. A signed transaction to an old address is stuck in the mempool. It’s supposed to be included in a block in thirty minutes. But fifteen minutes before that, a second transaction appears on the network—with the same signature but a different recipient. It pays a higher fee. Miners are rational: they include the more profitable one. The first transaction becomes history that never happened.

This isn’t a prediction or proof. It’s  a scenario that shows exactly where the fault line lies. To understand it, we need to answer a question rarely asked in popular texts:  what in the blockchain is protected by what cryptography and for how long?

Blockchain differs fundamentally from traditional cryptosystems. In a bank, a key can be revoked, reissued, and access restored through customer support. In blockchain,  the owner of the resource is the one who holds the private key . There is no central authority, no offline backup, and no recovery procedure. The ledger is always online and is both the asset and its description. Losing a key means the irreversible loss of the asset; compromising a key means irreversible theft. It is this rigid link between data and cryptosystem that makes blockchain particularly sensitive to advances in quantum technologies.

The key idea of ​​the original article:  Blockchain is not just a cryptosystem protecting an information asset. Blockchain  is  the asset itself. Therefore, any cryptographic weakness here translates directly into financial loss, without intermediaries and without the possibility of reversal.

This leads to a paradox worth keeping in mind for the rest of this text:  a more private system may be a less attractive target, but it’s not more resilient . Privacy increases the attacker’s uncertainty, not reduces the computational complexity of hacking. This distinction—between cryptography and attack economics—will form the core of our analysis.

2. Theme, central question and four hypotheses

⚙️ Detailed research option

The topic of the research is  to compare the quantum vulnerability of cryptographic subsystems of five public blockchains and to verify which conclusions about their relative strength remain valid under explicit assumptions.

The central question is:  which elements of blockchain systems actually become insecure when a quantum computer of sufficient size appears, and which conclusions about the relative vulnerabilities of Bitcoin, Ethereum, Litecoin, Monero, and ZCash hold up to scrutiny, and which are a consequence of the model rather than fact?

Four testable hypotheses

⚙️ Detailed research option

H1 · Captions are the main bottleneck

⚙️ Detailed research option

In all five systems, the critical vector is not hashing or consensus, but the elliptic curve digital signature scheme. The vulnerability is determined by the public key availability window.

H2 · Privacy changes the economy, not sustainability

⚙️ Detailed research option

Monero, ZCash, and Mimblewimble don’t improve cryptographic resistance to Shor, but they do increase target uncertainty. The “Monero is more secure” rating is only valid as an economic assessment.

H3 · Grover doesn’t provide a practical 51% attack in the near future

⚙️ Detailed research option

Quadratic gain can be compensated for by adjusting difficulty, except in systems with unique structures where the best quantum algorithm is known (ZCash/Equihash).

H4 · ZCash – Unique Emission Vector

⚙️ Detailed research option

If the private key of a zk-SNARK’s global public parameter is computable, tokens are issued “out of thin air,” invisible to the network. This is a fundamentally different risk than the theft of an individual transaction.

Refutation criteria

⚙️ Detailed research option

  • H1 is refuted if it turns out that for a typical transaction the public key is unavailable to an attacker for longer than the transaction lifetime, or if a quantum attack on a hash function turns out to be cheaper than an attack on a signature.
  • H2 is refuted if the cost of an attack per unit of asset value is independent of privacy: for example, an attacker can batch-break Pedersen commitments without increasing the cost per candidate.
  • H3 is refuted if the model shows that, under realistic parameters, a quantum miner outperforms a classical network faster than the difficulty recalculation is triggered.
  • H4 is refuted if resource assessment shows that recovering the global key requires error correction orders of magnitude greater than stealing a single signature—or if the trust ceremony guarantees the destruction of “toxic waste” in a verifiable manner.

Research Boundaries.  We do not estimate the exact date of the emergence of a cryptographically relevant quantum computer; we do not consider attacks on the network layer, wallets, or smart contracts per se; we do not design a new cryptosystem. We examine which statements about blockchains follow from the threat model and which from assumptions.

3. Method: How to distinguish evidence from prediction

⚙️ Detailed research option

The original paper considers two main vectors: attacks on  the signature scheme  using Shor’s algorithm and attacks on  the consensus mechanism  using Grover’s algorithm. For each blockchain, we additionally capture unique vectors that are not reducible to these two. Each assertion is labeled with a level of evidence:

LabelWhat does it mean?Example
A · provenA mathematical result or a reproducible experimentShor solves discrete logarithm in polynomial time on a quantum computer
B · ratedResource assessment under explicit assumptions485,550 qubits at 10 GHz give an attack on secp256k1 in ~30 minutes
C · forecastTrend extrapolation, scenario“By 2035, a computer that can break RSA-2048 will likely be available.”
D · qualitativelyArgument about the attractiveness of the target, without a numerical model“Private transactions are less attractive to attackers.”

A special disclaimer: in the PDF version of the original paper, the summary table of vulnerabilities is typographically damaged (symbols overlap). Therefore, below we  reconstruct  the summary from the section text rather than copying the table’s symbols. This doesn’t change the conclusions, but it honestly identifies the source of the uncertainty.

4. Mechanism: Shor, Grover, and the Special Fragility of Blockchain

⚙️ Detailed research option

4.1 Shor’s Algorithm: Periodicity Instead of Enumeration

⚙️ Detailed research option

Shor reduces factorization and discrete logarithm to the problem of finding the period of a function. A quantum circuit superimposes inputs, calculates the function, and, using interference, “assembles” the answer so that periodicity is manifested with high probability. Formally, for  nan -bit input, the time is  O(n³), and the memory is  O(n). For a classical computer, the discrete logarithm on an elliptic curve requires exponential time.

classical: ~O(2 n ) → quantum (Shor): O(n³) time, O(n) memory

A minimal example, without substituting an analogy.  Imagine you need to find the period of a repeating sequence, but you only see a “slice” of it. Classically, you’d try the shifts. Quantumly, you’d construct an interference pattern in which the incorrect periods cancel each other out, while the correct one is amplified. This works because the quantum state stores all variants simultaneously—but that’s precisely why you need a storage medium capable of holding a state the size of the input.

The limits of analogy.  “42 minutes at 10 MHz” isn’t a specification for a laboratory device. It’s an estimate for a machine that’s already capable of storing a state in quantum memory sufficient for the task at hand and has error correction. The real question isn’t the clock rate, but  how many physical qubits and what error per operation  are needed to produce one logical qubit.

4.2 Grover’s Algorithm: Quadratic Gain

⚙️ Detailed research option

Amplitude amplification allows us to find a solution in a space of  N elements per  O(√N) call. For blockchain, this means that any consensus based on solving a verifiable computational problem (PoW) can, in principle, achieve quadratic speedup. But it’s  quadratic , not exponential.

classical: N → quantum (Grover): √N

Boundaries.  The quadratic gain can be offset by increasing the problem difficulty: the network recalculates the target value, and the quantum miner’s advantage is partially “eaten away.” Furthermore, parallel Grover provides speedup only as the square root of the number of parallel machines, not linearly. Therefore, an attack on consensus is not a matter of “is the quantum computer faster?” but a complex resource race, decided by memory, errors, and economics.

4.3 Why is blockchain more fragile than traditional cryptosystems?

⚙️ Detailed research option

  • Key = active.  No recall, no recovery, no offline copy.
  • The public key is frequently revealed.  In the UTXO model, it becomes public when signing a transaction; in the account model, it becomes public upon the first outgoing transaction.
  • History is immutable.  Everything that’s been published remains accessible forever—convenient for “collect now, decipher later.”
  • A transaction is a race.  A transaction that has been signed but not yet included in a block creates a window in which an attacker can try to beat the owner.

5. Resource Model: What 42 Minutes Really Mean

⚙️ Detailed research option

The numbers from the original paper are useful as a scale, but dangerous as a forecast. Let’s summarize them in a table with explicit assumptions.

TaskClassical assessmentQuantum valuationKey assumption
RSA-2048~13.7 billion years on a 5 GHz CPU~42 minutes on a 10 MHz deviceThe device already maintains the desired state and correctly corrects errors
ECDSA secp256k1 (Bitcoin)Exponentially485,550 qubits at 10 GHz, ~30 minutesAggarwal et al.’s estimate; idealized error correction
51% attack on BitcoinASIC ~18 TH/s per deviceEarly boundary ~2028Forecast: ASIC improvements push back the date
51% attack on LitecoinNetwork ~320 TH/s, difficulty ~11.2 millionNeed a temp of ~2.4 THzCurrent network parameters; hashrate drop changes output
ZCash / Equihash (k-xor)Wagner’s algorithm O(2n/(1+⌊log₂k⌋))The algorithm of Grassi et al. O(2n/(2+⌊log₂k⌋))Quantum improvement in time and memory; parameters n, k are fixed

What these numbers don’t mean.  They don’t mean that “in 2035, any laptop with a quantum chip will steal your Bitcoin.” They mean that  if  a fault-tolerant machine with a sufficient number of logical qubits emerges,  the  cost of attacking a specific signature will become negligible compared to the value of the asset being protected. The gap between “physical qubits” and “logical qubits” is the main source of uncertainty.

The “2035” date in the paper is an extrapolation of trends in qubit count, gate accuracy, error correction, and fault tolerance, supported by expert consensus and official policy (NIST has launched the standardization of post-quantum cryptography). This is  a C forecast , not a schedule. The correct formulation is: “in the medium term, the probability is sufficient to plan for migration now.”

6. Five Cases: From Bitcoin to ZCash

⚙️ Detailed research option

To link sections of the original work into a coherent study, we use a single template:  signature → key availability window → consensus → unique vector → mitigation → level of evidence .

6.1 Bitcoin: The Base Case

⚙️ Detailed research option

Signature.  ECDSA on the secp256k1 curve. The signature consists of  R and  S:

S = k⁻¹ (SHA256(m) + dA R) mod p

During verification, the participant is required to reveal the public key associated with the address. This is the entry point for Shor.

Attack window.  Two fundamentally different cases. The first is  a mempool race : as soon as a broadcast transaction occurs, the public key is available; the attacker computes it  dA, signs a competing transaction, and tries to get into the block first. Estimate: 485,550 qubits at 10 GHz, ~30 minutes. The average confirmation time in Bitcoin often exceeds 30 minutes—the window is realistic. The second is  old P2PK addresses : in early transactions (including Hal Finney’s first broadcast, block 170), the public key was revealed forever. There is no time limit here: as soon as a machine appears, the key can be computed and the address emptied.

Consensus.  Hashcash on SHA-256. Grover provides quadratic speedup, but the network adjusts difficulty; ASIC technologies are pushing back the practical 51% attack. The authors indicate an early target of ~2028, but they themselves stipulate that hardware improvements are pushing it back.

Unique vector.  The Bitcoin Post-Quantum project (forked at block 555,000, December 22, 2018, XMSS scheme) does not protect the original chain—this is an important distinction between a “post-quantum fork” and “post-quantum Bitcoin.”

Mitigation:  Don’t reuse keys; migrate old P2PK; hybrid signatures; migrate to post-quantum schemes.  B · rated

6.2 Ethereum: Structural Key Reuse

⚙️ Detailed research option

Signature.  ECDSA; account model. As soon as an account sends its first outgoing transaction, its public key is known forever. This is qualitatively worse than the UTXO model: an attacker gains permanent access to the account, not a one-time window.

Consensus.  At the time of writing, Ethash (PoW on SHA-3/Keccak), with a plan to transition to Casper (PoS).  Post-publication clarification:  Ethereum transitioned to Proof-of-Stake in 2022 (The Merge). This removes the “quantum PoW miner” aspect from the article’s model, but does not remove the “revealed account public key” aspect. Moreover, PoS introduces its own challenges: validator key security, long-range attacks, and staking economics.

A unique vector.  Smart contracts and contract administrator keys; custodial contracts, in which a single key controls a large pool.

Mitigations.  Key rotation, account abstraction, post-quantum signatures, privileged key minimization.  B · rated

6.3. Litecoin and the Altcoin Family: Same ECDSA, Different Speed

⚙️ Detailed research option

Signature.  ECDSA secp256k1 is identical to Bitcoin’s. All Bitcoin forks and code clones inherit this vulnerability; the authors count 45 active forks.

Consensus.  Scrypt, a memory-intensive PoW. It hasn’t become ASIC-resistant in practice. The current hashrate is ~320 TH/s, difficulty ~11.2 million; a 51% attack requires a rate of ~2.4 THz. This makes an attack on consensus unlikely in the foreseeable future, but a drop in hashrate (for example, after a reward reduction) changes the picture.

Attack window.  Block length is ~2.5 minutes—shorter than Bitcoin’s, so the mempool race window is slightly smaller. This is  a minor  improvement: if a machine capable of attacking Bitcoin emerges, a small increase in clock speed will be enough to attack Litecoin as well.

Mitigation.  Same as Bitcoin; plus hashrate monitoring.  B · rated

6.4 Monero, Beam, and Grin: Privacy as an Economic Barrier

⚙️ Detailed research option

Caption:  EdDSA on the Ed25519 curve is a variant of ECDSA that also relies on discrete logarithm complexity. Caption:

S = (r + H(R, A, M) a) mod L

Additionally: ring signatures, stealth addresses, RingCT, Pedersen commitments, and range proofs (Bulletproofs). All of these also rely on the discrete logarithm.

Attack window.  The attacker doesn’t know the transfer amount or which of the ring inputs is the real one. To find the true key, several Pedersen commitments must be solved. Each solution is a separate run of Shor. This doesn’t make the attack impossible, but it increases the cost per unit of uncertainty.

Consensus.  CryptoNight v8 → RandomX (random program execution). At the time of writing, there is no known quantum advantage for RandomX—this is  the absence of a known attack , not a proof of security.

A unique vector.  Bulletproofs and Pedersen pledges are vulnerable to Shor; if a pledge is compromised, the transaction history is revealed. There may be no financial gain, but the value of the confidential information remains, creating a blackmail vector.

Beam and Grin.  Mimblewimble, blinding factor; same logic: obfuscation increases uncertainty, but not resistance.

Mitigation.  Post-quantum commitments and range proofs; transparent ceremonies; ring minimization.  D · qualitatively  +  B

6.5. ZCash: A Catastrophic Emission Vector

⚙️ Detailed research option

Signature.  EdDSA on Ed25519, hash BLAKE2b-256. As everywhere else, it relies on the discrete logarithm.

Consensus.  Equihash is a memory-intensive proof-of-work algorithm based on the generalized birthday problem (k-xor). Classically, it is solved by Wagner’s algorithm. Grassi et al. proposed a quantum algorithm with improved complexity:

Wagner: O(2 n/(1+⌊log₂k⌋) ) → quantum: O(2 n/(2+⌊log₂k⌋) )

This opens the way to a quantum 51% attack on consensus, which is more realistic than Bitcoin and Litecoin due to the structure of the problem.

A unique vector.  A zk-SNARK with a trusted ceremony. The global public parameter is designed such that the corresponding private key “should not exist.” If it is discovered (by Shor), the attacker gains the ability to create tokens in unlimited quantities. In a network with obfuscated amounts, this may go unnoticed. This isn’t theft—it’s  subverting the purpose of the ledger .

Mitigation.  Transparent post-quantum zk systems; ceremony auditing; trusted setup constraints.  B · rated  +  C

7. Comparison and rating – with reservations

⚙️ Detailed research option

BlockchainSignatures (Shor)Consensus (Grover)Main vectorResult
BitcoincriticalaverageMempool Race; Old P2PKsHigh
Ethereumcriticallow (PoS)Permanently exposed account keyHigh
LitecoincriticallowSame race, shorter windowHigh
Monerocriticalthere is no knownTheft + story revealAverage (economically)
ZCashcriticalincreasedEmission “out of nothing” + theftVery high

The original paper concludes:  Monero is the most stable  of those considered, followed by Bitcoin and Litecoin, then Ethereum, and then ZCash. This ranking is useful, but it should be read with two caveats.

Disclaimer 1.  “Monero is more secure” is a statement about  the attractiveness of the target , not about cryptographic strength. Once an attacker has chosen a specific transaction, the cost of breaking it is determined by Shor and ring size, not by privacy per se.

Disclaimer 2.  The ranking depends on the scenario. In the “one rich old address” scenario, Bitcoin may be worse than Monero; in the “mass history harvesting” scenario, private chains, on the contrary, provide the attacker with more hidden information, which will eventually become unreadable.

Research chain: how five cases come together to form one conclusion

⚙️ Detailed research option

  1. General model.  Shor breaks signatures; Grover provides a quadratic consensus gain.
  2. Bitcoin.  Base case: mempool window + permanently exposed old keys.
  3. Ethereum.  Structural key reuse turns a window into a permanent exposure.
  4. Litecoin.  The altcoin family inherits the same vulnerability; speed changes the window, but not the nature of the risk.
  5. Monero.  Privacy increases the cost of attack per unit of uncertainty, but does not eliminate Shor.
  6. ZCash.  A unique vector: not theft, but emission; a risk to the essence of the ledger, not just to the balance.
  7. Synthesis.  Signatures are the migration priority; consensus is the second priority; private chains require a separate post-quantum program.

8. Critical analysis: where the article is strong and where it is conditional

⚙️ Detailed research option

Strengths.  The work clearly distinguishes between two classes of algorithms, uses a common template for five systems, specifies a threat model, and does not conflate “transaction theft” with “account takeover.” This provides a good basis for comparative analysis.

What needs to be verified.

  1. Resource estimates.  Numbers like 485,550 qubits depend on the error correction model. Estimates in the literature differ by orders of magnitude.  How to verify:  reproduce the design with explicit error parameters and compare with alternative estimates.
  2. The date is 2035.  This is an extrapolation, not a law.  How to test it:  construct “optimistic/baseline/conservative” scenarios and show which conclusions are consistent.
  3. The “no PoW is Grover-proof” argument  is strong as a theoretical observation, but is weakened in practice by difficulty adjustments, memory-intensive functions, and parallel Grover constraints.  How to test this:  simulate the difficulty recalculation under a quantum miner.
  4. Ethereum before Merge.  Consensus analysis became obsolete after the transition to PoS.  How to test:  update the threat model for validators and long-range attacks.
  5. Privacy as “Security.”  Blending Cryptographic Strength and Attack Economics.  How to Test:  The Cost-Per-Candidate Model for Ring Signatures and Commitments.
  6. There’s no empirical data on the chain.  How much BTC is actually stored in P2PK and reused keys?  How to check:  chain analysis.
  7. Attacker rationality.  The model assumes that the attacker is willing to pay fees and take risks; in reality, there are sanctions, reputation, and KYC checks on the outputs.  How to test:  include the withdrawal cost and the risk of confiscation in the model.

9. A program of eight verifiable studies

⚙️ Detailed research option

Below are not “answers,” but questions that can be answered with the data. Each is phrased in a way that allows for refutation.

No.QuestionHypothesisMetricsDataWhat will refute
Q1What percentage of BTC is in P2PK and reused keys?H1: Old keys are the core of irreversible riskBTC, number of addresses, ageChain analysisThe share is negligible
Q2What is the probability of winning a mempool race at different speeds?A window of ~30 minutes is enoughP(win) from clock, commission, delaySimulation + historical dataP(win) is close to zero under realistic parameters
Q3Does privacy increase the cost of attack per unit cost?H2: Yes, but not durabilityCost per candidateModel + ring size distributionThe price does not depend on privacy
Q4Is the global key of zk-SNARK recoverable under realistic correction?H4: Yes, it is a unique vectorQubits, time, errorResource assessment + ceremony auditRequires orders of magnitude more resources than signature theft.
Q5Does recalculating the difficulty neutralize Grover’s advantage?H3: Basically yesQuantum Miner Network ShareAnalytical modelQuantum miner outpaces recalculation
Q6What is the price of post-quantum signatures in a block?Size is the main barrierBytes per transaction, commissionML-DSA / SLH-DSA / XMSSSize is irrelevant
Q7Does PoS remove the consensus vector?Removes the PoW vector, adds a keyValidator threat modelNetwork data after MergePoS doesn’t change the picture
Q8What is the “collect now – decrypt later” capacity for shielded pools?History accumulates riskVolume of hidden transactionsChain dataVolume is low or falling rapidly

10. Practice: what and in what order to change

⚙️ Detailed research option

The analysis doesn’t conclude that “all is lost,” but rather that a specific sequence of actions is needed. This aligns with NIST’s logic: FIPS 203 (ML-KEM), FIPS 204 (ML-DSA), and FIPS 205 (SLH-DSA) were approved in 2024; the NCCoE is leading the migration project.

  1. Signatures are the first priority.  ECDSA/EdDSA are being replaced by hybrid schemes (classical + post-quantum), then by fully post-quantum ones. XMSS is convenient, but statefulness complicates wallets.
  2. No key reuse.  This is already the norm for the UTXO model; for the account model, it requires account abstraction and rotation.
  3. Migrating old P2PKs.  This isn’t a technical issue, but a social one: owners must sign off on the transfer before the machine arrives.
  4. Consensus is the second priority.  PoS or memory-intensive schemes reduce, but do not eliminate, the advantage; difficulty adjustment is a mandatory element.
  5. Private chains are a separate program.  They require post-quantum commitments, range proofs, and transparent ceremonies.

The main practical conclusion is  that the blockchain doesn’t need to “rewrite everything urgently.” It needs  to start with signatures , because that’s where the gap between the cost of an attack and the value of the asset is greatest.

11. Open questions

⚙️ Detailed research option

  • What is the real cost of one logical qubit with error correction in the 2030s?
  • How much will quantum algorithms for k-xor improve after Grassi et al.?
  • Can a quantum computer speed up not only search but also   PoW verification ?
  • How do the economics of withdrawals (KYC, sanctions) change the attractiveness of an attack?
  • Do post-quantum zk-SNARKs with acceptable proof costs exist?
  • What happens to “assemble now, decode later” if the quantum advantage turns out to be more modest than expected?

12. Checklist before publication

⚙️ Detailed research option

  • Observations, evidence, estimates and forecasts are separated.
  • Every material statement has a source and level of evidence.
  • Physical qubits are not replaced by logical ones; simulation is replaced by experiment.
  • The advantage is compared under comparable conditions and accuracy.
  • There are no promises of a precise date for the hack or universal quantum acceleration.
  • Limitations and negative results are shown clearly.
  • The reader understands what new information the article provides and how to verify it.

13. Sources

⚙️ Detailed research option

  1. JJ Kearney, CA Perez-Delgado. Vulnerability of Blockchain Technologies to Quantum Attacks. arXiv:2105.01815v1. — The main text analyzed; all resource estimates and comparative conclusions for the five blockchains are cited there.
  2. PW Shor. Algorithms for Quantum Computation: Discrete Logarithms and Factoring. Proc. FOCS, 1994.
  3. LK Grover. Quantum Mechanics Helps in Searching for a Needle in a Haystack. Phys. Rev. Lett., 1997.
  4. D. Aggarwal, G. Brennen, T. Lee, M. Santha, M. Tomamichel. Quantum Attacks on Bitcoin, and How to Protect Against Them. Ledger, 2018. – Estimate 485,550 qubits / 30 minutes.
  5. NIST. FIPS 203, 204, 205: Post-quantum key establishment and signature standards (approved August 2024).
  6. NIST NCCoE. Migration to Post-Quantum Cryptography. — Practical Context for Migration.
  7. National Quantum Initiative. — Computing, networks, and sensors as different directions.
  8. S. Nakamoto. Bitcoin: A Peer-to-Peer Electronic Cash System. 2008.
  9. V. Buterin, V. Griffith. Casper the Friendly Finality Gadget. arXiv:1710.09437.
  10. Grassi et al. — a quantum algorithm for k-xor; cited under the ZCash section in [1].