wai-quantum: published architectures, re-estimated — the bicycle architecture, RSA-2048 and FeMoCo, each delta explained #665

Merged

dcharlot merged 6 commits from wai/quantum-published-architecture-check into main 2026-11-06 14:43:46 -05:00

Conversation 0 Commits 6 Files changed 13 +3446 -1

dcharlot commented 2026-11-06 03:17:33 -05:00

Owner

Copy link

Published fault-tolerant resource estimates, put through our estimators with their stated assumptions. For each quantity: the published figure, ours, the delta, and why (agrees, a different assumption, the source's model, the source's own figures disagreeing with each other or with its stated method, a limit of ours, or open). Nothing is fitted to the figures it is compared with, except one choice its row names (the Ising program's factory end).

The published figures live apart from the code, in wai/wai-quantum/tests/data/published_figures.tsv: each with where it is printed (section, table or figure; source files when quoted from the arXiv source with more digits than the PDF), as printed, its value and its rule. A row that agrees is held to the digits its source prints (exact, nearest printed unit, rounded up or down where the source rounds so, or within the range its fits' printed parameters allow); every other row carries a guard on ours to four digits. The page is wai/wai-quantum/docs/published-architectures.md; tests hold every row to its figure and the page to the code.

Commits

1. quantum_arch prices the 2021 surface-code machine (it could not express it). Its only clock waited a lattice-surgery beat per Toffoli: 57.3 ms for an addition over a 1,062-qubit piece (1,024 + the cost script's padding of 38) that the source runs at one 10 µs tick a Toffoli, 21.2 ms (its §2.11: "≈ 22 ms"). Adds the reaction-limited tick, double-access lookups, the two-level 15-to-1 / 8T-to-CCZ factory and the windowed factoring cost. Every row of arXiv:1905.09749's RSA table comes back to its printed digits. (Its commit message says 1,052; the review commit corrects it.)
2. quantum_bicycle (the estimators could not express a high-rate block code). A line of bivariate-bicycle modules, the architecture of arXiv:2506.03094.
3. published_architectures, the page, the example, the CI step and the changelog entry.
4. The review's fixes (below).

The headline rows

Quantity	Source	Published	Ours	Delta	Why
Table 1 / Fig. 10(a) qubit counts	arXiv:2506.03094	288, 90, 22, … 121 … 6,886	the same	0	agrees, exactly
Table 10's 24 fitted rates, its 7 failure probabilities, §A.10's two bounds	arXiv:2506.03094	e.g. 1.6·10⁻⁹, 3.33·10⁴, 7.32·10⁻⁴	e.g. 1.61e-9, 3.331e4, 7.322e-4		agrees, within the range Table 6's printed parameters allow
capability, gross 5k, p = 10⁻³	Fig. 10(a), source files	13.5958	13.65	+0.42%	assumption: ours divides P_I and P_U by C as §A.7 says (+0.30%); undivided, ours is 13.61 (+0.12%, their sampling and fit, Fig. 17)
capability, two-gross rows	Fig. 10(a), source files	3.4712·10⁶ …	3.478e6 …	+0.2–0.3%	their model: sampling and a fitted line against our expected count
capability, gross p = 10⁻⁴ (5k / 50k)	Fig. 10(a), source files	2.2209·10⁵ / 3.7033·10⁴	4.701e5 / 4.064e4	×2.12 / +9.7%	open: its rows need 8.3·10⁻⁷ a rotation whatever the size, against its factory's 4·10⁻⁸; its released compiler (bicycle_compiler 0.1.0) would settle it
timesteps per T	Fig. 10(a), source files	3367.15 …	3191 …	−7.1% to +0.05%	our limit: no scheduling
mean in-module measurements per Pauli (4)	Fig. 9	18.48, 18.58, 18.55, 18.67	18.526, 18.482, 18.593, 18.561	≤ 0.6%	their numbers: bicycle_cliffords 0.1.0 src/measurement.rs takes Z̄ by A⁻¹ where a CNOT circuit takes it by A⁻ᵀ; so taken, our search gives all four to every printed digit (0.2.1 takes X̄ by A⁻¹, Z̄ by Aᵀ)
native rotations	§3.2	510	245	×0.48	their numbers: 510 distinct R ≠ I over all natives (284 with Q = I, 19 also with Q ≠ I); that it counted these is our inference
idle and shift rates per cycle	§A.7 against Table 2	10^−8.8	1.61e-10	×0.10	their numbers: the tables print the circuit's failure undivided
memory per cycle	arXiv:2308.07915 Table 1	2·10⁻⁷	1.61e-10	×8.0e-4	open: about 1,250× between the two papers, not explained here
Ising T gates	Table 9, source files	15,542,400	14,665,600	−5.6%	their model: pygridsynth 1.0.1 (the version it cites) gives 86, 80, 84, 82 a rotation and its total exactly; ours 79, 80, 81, 84
Ising inter-module measurements	Table 9	9.4680·10⁵	946,800	0	assumption: the factory after the last module; before the first, 781,200; the paper prints neither
RSA-2048 (2021) re-run: qubits, hours, Toffolis	arXiv:1905.09749 Tables 3, 1	20 M, 5.1, 2.7 bn	19,248,768, 5.05, 2.698e9		agrees (its script rounds up to two digits)
its title's hours	title	8	7.31		agrees: expected across its 30.98% retries, taken up to one digit
the same counts in our estimate job: hours	Table 3	5.1	60.71	×11.9	our limit: the job prices the planar architecture with no reaction-limited clock
FeMoCo: qubits, days, Toffoli rate	arXiv:2011.03494 Fig. 10, §IV.C.4	≈ 4·10⁶, 3, 25 kHz	3,907,584, 3.01, 25.8 kHz		agrees (25 kHz is 25.8 printed down)
RSA-2048 (2025): qubits, hours a shot, days	arXiv:2505.15917 §3.2	897,864, 12.07, 4.96	958,480, 10.58, 4.31		their numbers
planar model's source	arXiv:2211.07629 Appendix F	13, 18, 8.72M, 17 h 43 min	13, 18, 8,716,258, 17.72 h		agrees

The review's fixes (commit 4)

1. "Its compiler is not released" was false: bicycle_compiler, bicycle_cliffords and bicycle_common 0.1.0 are on crates.io. The compile-mean rows cite bicycle_cliffords 0.1.0 src/measurement.rs (A on X̄, A⁻¹ on Z̄; 0.2.1: A⁻¹ on X̄, Aᵀ on Z̄). The two gross p = 10⁻⁴ rows stay open, naming the compiler as what would settle them. No third-party code was downloaded or run.
2. "Agrees" was 6% for every row. Each row's rule now comes from its printed digits or rounding rule, in the pinned figures file; non-agree rows carry guards.
3. The six surface-code timesteps are Fig. 10(a)'s printed figures.
4. The capability rows' reason now says ours divides P_I and P_U by C, and how much of the gap that is; a row beside each gives ours undivided.
5. Fig. 17, not Fig. 16.
6. "Ours" holds only ours.
7. The Ising factory's end is a stated rule (Machine::links), and the row is an assumption that says it was chosen to match.
8. The 510 wording: 284 with Q = I, 19 also with Q ≠ I, 265 only with Q = I; the count's meaning is our inference.
9. No more digits than the PDF unless "source files".
10. The expected hours are one computation (7.31).
11. The arXiv:2308.07915 rows are open.
12. The 2021 piece is 1,062 qubits (padding 38), not 1,052.
13. The section on the 2029 machine, its sources and its allow-name are gone; the synthesiser row has no shortfall line (its fix is #671, stacked on this one). The first fix left the product's name in a test, as a string that must appear in no row; 3e78384b5 takes it out. The test keeps "press release" and "announced", and putting "press release" back in one pinned source still fails it.

What this does not do

• The estimate job has no reaction-limited clock (the ×11.9); quantum_arch has it. #673, stacked on this one, brings it into the job.
• The bicycle machine is not scheduled, and repeated in-module measurements are not removed.
• The bicycle rates are the paper's; no decoder here has run its instructions.
• Two rows are open (the gross capability at 10⁻⁴) and four (arXiv:2308.07915).

Checks run (build host, Linux x86_64)

• Release, on head 3e78384b5's tree, Linux aarch64: quantum_arch:: quantum_bicycle:: published_architectures::, 40 passed, 0 failed.
• Release, on head fedcc0976: cargo test --release --features full --lib -- quantum_arch:: quantum_bicycle:: published_architectures:: engine::estimate quantum_resource: 74 passed, 0 failed, 0 ignored, the page test among them.
• Debug, the three modules: 35 passed, 5 ignored (the release-only searches), 0 failed.
• RUSTFLAGS=-D warnings cargo check --no-default-features with quantum_arch, quantum_arch,quantum_engine and quantum_engine: clean.
• The review's mutation (the planar source's 8.72M re-pinned as 9.00M, the page regenerated) now fails every_row_holds_to_its_published_figure; it passed every test before.
• Public-name (the pinned figures file included), comparative-prose, vanity, spec-reference and changelog checks: clean. No dependency changed.

On this PR's first CI run (head 4506ada5f), the 14 red legs were macOS: their logs stop in compilation with the runner's sccache server failing ("error reading compile response from server"), not a test.

🤖 Generated with Claude Code

Published fault-tolerant resource estimates, put through our estimators with their stated assumptions. For each quantity: the published figure, ours, the delta, and why (agrees, a different assumption, the source's model, the source's own figures disagreeing with each other or with its stated method, a limit of ours, or open). Nothing is fitted to the figures it is compared with, except one choice its row names (the Ising program's factory end). The published figures live apart from the code, in `wai/wai-quantum/tests/data/published_figures.tsv`: each with where it is printed (section, table or figure; `source files` when quoted from the arXiv source with more digits than the PDF), as printed, its value and its rule. A row that agrees is held to the digits its source prints (exact, nearest printed unit, rounded up or down where the source rounds so, or within the range its fits' printed parameters allow); every other row carries a guard on ours to four digits. The page is `wai/wai-quantum/docs/published-architectures.md`; tests hold every row to its figure and the page to the code. ## Commits 1. **`quantum_arch` prices the 2021 surface-code machine** (it could not express it). Its only clock waited a lattice-surgery beat per Toffoli: 57.3 ms for an addition over a 1,062-qubit piece (1,024 + the cost script's padding of 38) that the source runs at one 10 µs tick a Toffoli, 21.2 ms (its §2.11: "≈ 22 ms"). Adds the reaction-limited tick, double-access lookups, the two-level 15-to-1 / 8T-to-CCZ factory and the windowed factoring cost. Every row of arXiv:1905.09749's RSA table comes back to its printed digits. (Its commit message says 1,052; the review commit corrects it.) 2. **`quantum_bicycle`** (the estimators could not express a high-rate block code). A line of bivariate-bicycle modules, the architecture of arXiv:2506.03094. 3. **`published_architectures`**, the page, the example, the CI step and the changelog entry. 4. **The review's fixes** (below). ## The headline rows | Quantity | Source | Published | Ours | Delta | Why | |---|---|---:|---:|---:|---| | Table 1 / Fig. 10(a) qubit counts | arXiv:2506.03094 | 288, 90, 22, … 121 … 6,886 | the same | 0 | agrees, exactly | | Table 10's 24 fitted rates, its 7 failure probabilities, §A.10's two bounds | arXiv:2506.03094 | e.g. 1.6·10⁻⁹, 3.33·10⁴, 7.32·10⁻⁴ | e.g. 1.61e-9, 3.331e4, 7.322e-4 | | agrees, within the range Table 6's printed parameters allow | | capability, gross 5k, p = 10⁻³ | Fig. 10(a), source files | 13.5958 | 13.65 | +0.42% | assumption: ours divides P_I and P_U by C as §A.7 says (+0.30%); undivided, ours is 13.61 (+0.12%, their sampling and fit, Fig. 17) | | capability, two-gross rows | Fig. 10(a), source files | 3.4712·10⁶ … | 3.478e6 … | +0.2–0.3% | their model: sampling and a fitted line against our expected count | | capability, gross p = 10⁻⁴ (5k / 50k) | Fig. 10(a), source files | 2.2209·10⁵ / 3.7033·10⁴ | 4.701e5 / 4.064e4 | ×2.12 / +9.7% | **open**: its rows need 8.3·10⁻⁷ a rotation whatever the size, against its factory's 4·10⁻⁸; its released compiler (bicycle_compiler 0.1.0) would settle it | | timesteps per T | Fig. 10(a), source files | 3367.15 … | 3191 … | −7.1% to +0.05% | our limit: no scheduling | | mean in-module measurements per Pauli (4) | Fig. 9 | 18.48, 18.58, 18.55, 18.67 | 18.526, 18.482, 18.593, 18.561 | ≤ 0.6% | their numbers: bicycle_cliffords 0.1.0 `src/measurement.rs` takes Z̄ by A⁻¹ where a CNOT circuit takes it by A⁻ᵀ; so taken, our search gives all four to every printed digit (0.2.1 takes X̄ by A⁻¹, Z̄ by Aᵀ) | | native rotations | §3.2 | 510 | 245 | ×0.48 | their numbers: 510 distinct R ≠ I over all natives (284 with Q = I, 19 also with Q ≠ I); that it counted these is our inference | | idle and shift rates per cycle | §A.7 against Table 2 | 10^−8.8 | 1.61e-10 | ×0.10 | their numbers: the tables print the circuit's failure undivided | | memory per cycle | arXiv:2308.07915 Table 1 | 2·10⁻⁷ | 1.61e-10 | ×8.0e-4 | **open**: about 1,250× between the two papers, not explained here | | Ising T gates | Table 9, source files | 15,542,400 | 14,665,600 | −5.6% | their model: pygridsynth 1.0.1 (the version it cites) gives 86, 80, 84, 82 a rotation and its total exactly; ours 79, 80, 81, 84 | | Ising inter-module measurements | Table 9 | 9.4680·10⁵ | 946,800 | 0 | assumption: the factory after the last module; before the first, 781,200; the paper prints neither | | RSA-2048 (2021) re-run: qubits, hours, Toffolis | arXiv:1905.09749 Tables 3, 1 | 20 M, 5.1, 2.7 bn | 19,248,768, 5.05, 2.698e9 | | agrees (its script rounds up to two digits) | | its title's hours | title | 8 | 7.31 | | agrees: expected across its 30.98% retries, taken up to one digit | | the same counts in our estimate job: hours | Table 3 | 5.1 | 60.71 | ×11.9 | our limit: the job prices the planar architecture with no reaction-limited clock | | FeMoCo: qubits, days, Toffoli rate | arXiv:2011.03494 Fig. 10, §IV.C.4 | ≈ 4·10⁶, 3, 25 kHz | 3,907,584, 3.01, 25.8 kHz | | agrees (25 kHz is 25.8 printed down) | | RSA-2048 (2025): qubits, hours a shot, days | arXiv:2505.15917 §3.2 | 897,864, 12.07, 4.96 | 958,480, 10.58, 4.31 | | their numbers | | planar model's source | arXiv:2211.07629 Appendix F | 13, 18, 8.72M, 17 h 43 min | 13, 18, 8,716,258, 17.72 h | | agrees | ## The review's fixes (commit 4) 1. "Its compiler is not released" was false: bicycle_compiler, bicycle_cliffords and bicycle_common 0.1.0 are on crates.io. The compile-mean rows cite bicycle_cliffords 0.1.0 `src/measurement.rs` (A on X̄, A⁻¹ on Z̄; 0.2.1: A⁻¹ on X̄, Aᵀ on Z̄). The two gross p = 10⁻⁴ rows stay open, naming the compiler as what would settle them. No third-party code was downloaded or run. 2. "Agrees" was 6% for every row. Each row's rule now comes from its printed digits or rounding rule, in the pinned figures file; non-agree rows carry guards. 3. The six surface-code timesteps are Fig. 10(a)'s printed figures. 4. The capability rows' reason now says ours divides P_I and P_U by C, and how much of the gap that is; a row beside each gives ours undivided. 5. Fig. 17, not Fig. 16. 6. "Ours" holds only ours. 7. The Ising factory's end is a stated rule (`Machine::links`), and the row is an assumption that says it was chosen to match. 8. The 510 wording: 284 with Q = I, 19 also with Q ≠ I, 265 only with Q = I; the count's meaning is our inference. 9. No more digits than the PDF unless "source files". 10. The expected hours are one computation (7.31). 11. The arXiv:2308.07915 rows are open. 12. The 2021 piece is 1,062 qubits (padding 38), not 1,052. 13. The section on the 2029 machine, its sources and its allow-name are gone; the synthesiser row has no shortfall line (its fix is #671, stacked on this one). The first fix left the product's name in a test, as a string that must appear in no row; `3e78384b5` takes it out. The test keeps "press release" and "announced", and putting "press release" back in one pinned source still fails it. ## What this does not do - The estimate job has no reaction-limited clock (the ×11.9); `quantum_arch` has it. #673, stacked on this one, brings it into the job. - The bicycle machine is not scheduled, and repeated in-module measurements are not removed. - The bicycle rates are the paper's; no decoder here has run its instructions. - Two rows are open (the gross capability at 10⁻⁴) and four (arXiv:2308.07915). ## Checks run (build host, Linux x86_64) - Release, on head `3e78384b5`'s tree, Linux aarch64: `quantum_arch:: quantum_bicycle:: published_architectures::`, 40 passed, 0 failed. - Release, on head `fedcc0976`: `cargo test --release --features full --lib -- quantum_arch:: quantum_bicycle:: published_architectures:: engine::estimate quantum_resource`: 74 passed, 0 failed, 0 ignored, the page test among them. - Debug, the three modules: 35 passed, 5 ignored (the release-only searches), 0 failed. - `RUSTFLAGS=-D warnings cargo check --no-default-features` with `quantum_arch`, `quantum_arch,quantum_engine` and `quantum_engine`: clean. - The review's mutation (the planar source's 8.72M re-pinned as 9.00M, the page regenerated) now fails `every_row_holds_to_its_published_figure`; it passed every test before. - Public-name (the pinned figures file included), comparative-prose, vanity, spec-reference and changelog checks: clean. No dependency changed. On this PR's first CI run (head `4506ada5f`), the 14 red legs were macOS: their logs stop in compilation with the runner's sccache server failing ("error reading compile response from server"), not a test. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

dcharlot added 4 commits 2026-11-06 03:17:33 -05:00

wai-quantum: quantum_arch prices the 2021 surface-code machine ... 475f71f28d

The architecture estimator could not express the machine of
arXiv:1905.09749 (Quantum 5, 433), so its figures could not be checked.
Its only clock was the step clock: every Toffoli waits a lattice-surgery
beat of d cycles. That prices the source's 1,052-bit addition at
56.7 ms, where the source teleports each Toffoli's Clifford fixups away
from the data and runs it at one 10 µs tick: 22 ms (its §2.11). It also
had no two-level factory, which that machine and arXiv:2011.03494 use.

Added, each held by a test against the source's printed figures:

- Timing::tick_us and Timing::teleported_add_us: a Toffoli at the slower
  of the reaction time and the CCZ supply, not at the beat. The test
  holds the step clock's 56.7 ms beside the tick's 21.03 ms.
- Timing::double_access_lookup_us and its uncomputation: lookups through
  access hallways on both sides of each row, half a beat an entry.
- patch_cycle_error and DistilledCczFactory: 15-to-1 feeding 8T-to-CCZ,
  with the footprint and error rule of the source's cost script.
- WindowedFactoringParams and windowed_factoring_estimate: the source's
  cost script, re-run. Every row of its RSA table comes back to the
  digits it prints, which it takes up to two.
- The FeMoCo estimate of arXiv:2011.03494 on the same factory and clock:
  four factories at distances 19 and 31 tick at 38.75 µs, and 6.7·10⁹
  Toffolis take 3 days.

The module documents where the 2021 source disagrees with itself: its
hand-worked §2.14 rounds as it goes (23 M where its script gives
19.25 M), and its §2.12 reads 6.4% where its script's formula gives 0.6%.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

wai-quantum: quantum_bicycle, a machine of bivariate-bicycle code modules ... 84ab31111c

The estimators priced one logical qubit per surface-code tile and could
not express a high-rate block code: twelve logical qubits in one code
block, reached by a small instruction set instead of lattice surgery on
a tile. The machine announced for 2029 (200 logical qubits, 10⁸
operations) is built on such codes, on the bicycle architecture of
arXiv:2506.03094, so it could not be checked.

quantum_bicycle prices that architecture from the parts the paper
publishes:

- the gross [[144,12,12]] and two-gross [[288,12,18]] codes from their
  printed polynomials and logical basis (tests: the basis is dual, every
  operator commutes with every check, k = 12, and a shift acts as the
  printed matrices);
- the native measurements and, by breadth-first search over the 4¹¹ − 1
  Paulis on a module's data qubits, the in-module cost of each;
- the instruction rates from the paper's failure-spectrum fits (all 28
  printed rates come back), and its assumed rates, marked as assumed;
- a machine of M modules and a factory, q = M(c + u + a) − a + a′ + f
  (every count of its capability table comes back), its capability at a
  failure budget, and the error rate a program needs.

Where the paper disagrees with itself or with the method it states, the
module says so, and a test holds each:

- Its four mean compile costs (Fig. 9) come back to every printed digit
  when a shift's action on Z̄ is taken as A⁻¹ with A on both halves
  (ShiftAction::SourceTables). A CNOT circuit |x⟩ → |Ax⟩ takes Z̄ by A⁻ᵀ,
  which is how the operators move; ours follows the operators. On its
  printed basis a shift acts as A on the first half and Aᵀ on the second.
- Its idle and shift rates are the circuit's failure undivided, where its
  text divides by the circuit's cycles.
- Its native-rotation count of 510 includes the 265 with the identity on
  the pivot, which its definition excludes.

One difference stays open: its gross capability at p = 10⁻⁴ needs a
per-rotation failure of 8.3·10⁻⁷ that does not depend on the machine's
size, against its factory's printed 4·10⁻⁸.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

wai-quantum: published architectures, re-estimated ... 54b331394a

published_architectures feeds the stated assumptions of published
resource estimates into the estimators (quantum_arch, quantum_bicycle
and the estimate job) and prints, for each quantity, the published
figure, ours, the delta and why: agrees, a different assumption, the
source's model, the source's own figures disagreeing with each other or
with its stated method, a limit of ours, or open. Every row names where
its figure is printed; figures with more digits than a PDF prints are
quoted from the paper's arXiv source files and say so, and the one
secondary figure is marked secondary.

The targets:
- the machine announced for 2029, "100 million quantum operations using
  200 logical qubits" (press release of 2025-06-10), priced on the
  bicycle architecture of arXiv:2506.03094, the paper the release links;
- that paper's own figures: its parts, rates, compile, capabilities and
  10 × 10 Ising program;
- RSA-2048 on the surface code, arXiv:1905.09749 (Quantum 5, 433);
- FeMoCo, arXiv:2011.03494 (PRX Quantum 2, 030305);
- RSA-2048 on yoked surface codes, arXiv:2505.15917;
- the planar model's own source, arXiv:2211.07629.

Of 197 rows, 110 agree. Two are open: the source's gross capability at
p = 10⁻⁴ needs a per-rotation failure of 8.3·10⁻⁷ that its tables do
not account for. One row records a shortfall of ours that the check
found and this change does not fix: on one of the Ising program's four
angles our synthesiser takes 84 T gates where pygridsynth 1.0.1 takes
82, at the same denominator exponent.

docs/published-architectures.md holds the rows, and
the_page_is_what_the_estimators_give holds the page to the code; the
example prints them. The resource-estimation workflow runs the three
modules' tests in a release build, requires seven by name, and runs the
example.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

Merge origin/main into wai/quantum-published-architecture-check ... 4506ada5f7

Co-Authored-By: Claude Opus 5.5 <[email protected]>

dcharlot added 1 commit 2026-11-06 08:00:39 -05:00

wai-quantum: the published-architecture check, as its review asked ... fedcc09763

Every fix with a test that fails on the code before it.

Figures and rules
- The published figures move out of the code into
  tests/data/published_figures.tsv: each with where it is printed, as
  printed, its value and its rule, so regenerating the page cannot move
  them. A row that agrees is held to the digits its source prints:
  exact, the nearest printed unit, rounded up or down where the source
  rounds so, or, for a figure ours takes from the source's fits, the
  range its fits' printed parameters (ln f0 and gamma, two decimals)
  allow. It was 6% for every row: 8.72M re-pinned as 9.00M passed. Every
  other numeric row carries a guard, ours to four digits.
  (every_row_holds_to_its_published_figure,
  a_rule_holds_to_the_printed_digits_and_no_further)
- Fig. 10(a)'s surface-code timesteps are its printed 381, 2,269, 115,
  115, 509 and 509, not the formula compared with itself.
- Figures carry no more digits than the PDF prints (7.32e-4, 2.09e-5,
  8.2e-21, Table 10 to its printed two and three digits) unless the row
  says "source files". The Toffoli count of arXiv:1905.09749 is its
  Table 1's, not Table 3's.

Reasons and classes
- The authors' compiler is released (bicycle_compiler, bicycle_cliffords
  and bicycle_common 0.1.0 on crates.io, 2025-07-31). The compile-mean
  rows now cite bicycle_cliffords 0.1.0 src/measurement.rs, which takes
  a shift as A on X and A^-1 on Z (version 0.2.1: A^-1 on X, A^T on Z),
  for what was an inference. The two gross rows at p = 1e-4 stay open,
  with that compiler named as what would settle them; this check runs
  no third-party code.
- The capability rows said the rates and the sum were the same. Ours
  divides P_I and P_U by C (Section A.7's text): a third of a point of
  the gross 5k p = 1e-3 gap is that choice. That row is now an
  assumption, with a row beside each capability at the undivided rates.
  (the_division_by_c_is_a_third_of_a_point_of_the_gross_capability)
- The random-circuit fit is Fig. 17, not Fig. 16.
- The "Ours" column holds only ours: the Ising T-count-as-stated row
  shows our 14,665,600; the FeMoCo Table III against layout rows are the
  paper's own two figures, ours empty, their numbers.
- The Ising inter-module count depends on the factory's end of the line
  (946,800 after the last module, 781,200 before the first), which the
  paper does not print. Machine::links states the rule, a test computes
  both, and the row is an assumption that says it was chosen to match.
- The 510: over all native measurements there are 510 distinct R != I,
  284 with Q = I, 19 of those also with Q != I. That the paper counted
  these is our inference from the matching count, and the text says so
  (native_rotation_census).
- The arXiv:2308.07915 rows (about 1,250 times the later fit) are open.
- The expected hours of arXiv:1905.09749's title are one computation,
  WindowedEstimate::expected_hours: 7.31 in the row and the module.

Facts
- The 2021 addition's piece is 1,024 + 38 = 1,062 qubits (the cost
  script's padding; its text's 3 lg n + 10 is 43), not 1,052:
  WindowedFactoringParams::padding and piece_qubits, and the test.

Public text
- The section on the machine announced for 2029, its sources and its
  allow-name are gone, with every verdict on that announcement. The
  rows that check arXiv:2506.03094 itself stay.
- The synthesiser row says what the counts are, without the shortfall
  line; its fix follows separately.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

dcharlot referenced this pull request 2026-11-06 08:17:25 -05:00

wai-quantum: a rotation's Clifford+T word takes the cheapest unit of the solver's solution, never more T gates than the reference implementation #671

dcharlot referenced this pull request 2026-11-06 08:40:09 -05:00

wai-quantum: the estimate job on the reaction-limited clock, one tick rule shared with quantum_arch, the step clock's answers and receipts unchanged #673

dcharlot changed title from wai-quantum: published architectures, re-estimated — the 2029 machine, the bicycle architecture, RSA-2048 and FeMoCo, each delta explained to wai-quantum: published architectures, re-estimated — the bicycle architecture, RSA-2048 and FeMoCo, each delta explained 2026-11-06 09:50:10 -05:00

dcharlot added 1 commit 2026-11-06 10:22:42 -05:00

wai-quantum: the guard on the removed section names no product ... 3e78384b5f

The review asked for the section on the announced machine to go, and
the fix for that left the product's name in a test, as a string that
must appear in no row. The name was a bare literal in a public file, and
it guarded nothing the other strings did not: the removed rows' sources
read "press release", and their text "announced".

The test keeps those two. Put "press release" back in one pinned source
and it still fails, naming that string.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

dcharlot referenced this pull request from a commit 2026-11-06 10:22:43 -05:00

merge the guard fix from #665

dcharlot referenced this pull request from a commit 2026-11-06 10:22:45 -05:00

merge the guard fix from #665

dcharlot merged commit 092568754b into main 2026-11-06 14:43:46 -05:00

dcharlot referenced this pull request from a commit 2026-11-06 14:43:49 -05:00

Merge pull request 'wai-quantum: published architectures, re-estimated — the bicycle architecture, RSA-2048 and FeMoCo, each delta explained' (#665) from wai/quantum-published-architecture-check into main

Sign in to join this conversation.

Reviewers

No reviewers

Labels

Clear labels

No items

No labels

Milestone

Clear milestone

No items

No milestone

Projects

Clear projects

No items

No project

Assignees

Clear assignees

No assignees

1 participant

Notifications

Subscribe

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference

Transaction-Science/open-standards!665

No description provided.