Cubrim — Global Addresser

The honest limit, stated first

A fixed short code can distinguish only a finite number of states, while the space of possible files grows exponentially with length — so a short address alone can never uniquely denote every long sequence. An address is meaningful only together with a catalog where the object already exists. And any honest result must charge the full cost: the addresses, the metadata, the residual data — and the shared matrix itself, which is only worth its size when amortized across many files and devices. Cubrim-2 therefore does not promise to “compress any file into 16 bytes”. It maps where global addressing genuinely beats local compression — and records, just as openly, where it structurally cannot.

How much space would ALL possible matrices take

We count in bits. An N-dimensional cube is a visual arrangement of a bit sequence: a 4×4×4 cube is 64 bits laid out in three dimensions. Dimensionality (1D/2D/3D/4D) is a way to arrange the same bits, not a different amount of them. Hence the number of all possible matrices of length B bits is 2^B and depends only on B: a 64-bit 1D string and a 4×4×4 cube share the same 2^64 ≈ 1.8×10^19 states.

B, bits	cube example	all possible = 2^B	store them all (bytes)	fits in 20,000×1 TB?
8	2×2×2	2.56×10²	2.56×10² B	yes
27	3×3×3	1.34×10⁸	4.53×10⁸ B	yes
51	—(порог/threshold)	2.25×10¹⁵	1.44×10¹⁶ B	yes
64	4×4×4	1.84×10¹⁹	1.48×10²⁰ B	no
125	5×5×5	4.25×10³⁷	6.65×10³⁸ B	no
256	4×4×4×4 (4D)	1.16×10⁷⁷	3.71×10⁷⁸ B	no
1000	10×10×10	1.07×10³⁰¹	1.34×10³⁰³ B	no
4096	8×8×8×8 (4D)	1.04×10¹²³³	5.35×10¹²³⁵ B	no

Capacity yardstick: 20,000 disks of 1 TB = 2×10^16 bytes = 1.6×10^17 bits. The full-enumeration threshold is B = 51 bits: the last length at which all 2^B matrices (with their contents, 2^B×B bits = 1.44×10^16 bytes) still fit; at B = 52 they no longer do (2.93×10^16 bytes). A 4×4×4 cube (64 bits): just the 64-bit addresses of all 2^64 matrices would take 1.48×10^20 bytes ≈ 147.6 EB — four orders of magnitude beyond the yardstick. At B = 256 there are 1.16×10^77 matrices — still slightly fewer than the atoms in the observable universe (~10^80); the “more than atoms” line is crossed at B ≈ 266. Beyond that, super-exponential growth with no physical storage prospect whatsoever.

The honest conclusion: “all possible matrices” cannot be enumerated — 2^B diverges super-exponentially at lengths below a single machine word. So the Addresser stores not all possible blocks but only the ones ACTUALLY ENCOUNTERED (CAS/deduplication): their number is bounded by the volume of real data and obeys the track’s measured laws — 89.9% of blocks in an uncurated matrix occur exactly once (AH-19), and an honest 16–64 bytes per object is achievable only on an exact match against the catalog (AH-05). This is the same fundamental limit as the block above: an address is meaningful only together with a catalog where the object already exists.

On dimensionality: 2^B does not depend on it, but dimensionality changes which bits land in one cube when REAL data is tiled — and therefore how many unique cubes occur in practice. That is measured by a scan (part B of the experiment), not by a formula.

Formulas: number of matrices = 2^B; storing the contents = 2^B × B bits; the ordinal address of one matrix = B bits (2^B states). All numbers are computed with exact integer arithmetic (python), not estimates.

How much space the ACTUALLY encountered matrices need (scan experiment)

A real CROSS-DEVICE corpus of 13.48 GB (a union of three hosts: arcana-devs 9.06 + arcana-www 3.38 + arcana-prod 1.04 GB; the hosts exchanged only cube hashes) was tiled into cubes of B = 4096 bits (512 bytes) in four layouts of the same length: a 1D string of 4096, 2D 64×64, 3D 16×16×16, 4D 8×8×8×8. A layout groups DIFFERENT bits of the file into one cube (strided tiling with a 4096-byte row), so the number of unique cubes on real data differs — even though the number of possible ones (2^4096) is identical.

layout	cubes scanned	unique	% unique	cross-host overlap (cubes)	store the unique
1D 4096	25 890 520	14 965 607	57.80%	1 847 617	7.14 GiB
2D 64×64	19 368 448	12 077 059	62.35%	1 260 924	5.76 GiB
3D 16×16×16	13 959 168	8 020 728	57.46%	943 354	3.82 GiB
4D 8×8×8×8	10 518 528	5 578 363	53.03%	768 085	2.66 GiB

There is NO saturation: the number of unique cubes grows almost linearly to the very end of the scan (devs-slice curve: 1D 10.9M unique at 7.3 GB → 12.3M at 8.3 GB) — real data at this cube length hardly “runs out”. On the devs slice, 72–82% of unique cubes were seen exactly once — the same measured law as AH-19 (89.9% on CDC chunks): an uncurated matrix is mostly dead weight. Cross-host overlap is real (0.77–1.85M cubes are shared between hosts), and it is what lowers the union corpus unique share to 53–62%.

Disks for the scan itself: 2.7–7.1 GiB of unique matrices per 13.48 GB of cross-device data — a fraction of one 1 TB disk. Extrapolating to the world’s stored data (on the order of 10 ZB = 10^22 bytes; order-of-magnitude per IDC Global DataSphere reports): with the measured 53–62% unique share on the cross-device corpus and NO saturation, the matrices would take ~5.3–6.2 ZB — billions of 1 TB disks, i.e. the same order as the data itself. THE EXTRAPOLATION IS LINEAR AND MARKED AS AN ASSUMPTION: the world corpus differs in composition, and saturation at larger volumes is not excluded — it was simply not observed in the scan.

Conclusions: (1) even storing only the ENCOUNTERED matrices at world scale is the same order as the data itself — the Addresser’s win lives not in a “warehouse of all cubes” but in deduplicating the repeated share (38–47% on the union corpus) and in r≥2 curation (AH-19); (2) the optimal layout by the “fewer unique” criterion is 4D (53.0% unique vs 62.4% for 2D on the union corpus) — multidimensional grouping does gather repeating bits more often, with a moderate margin (~6–9 pp); (3) the comparison is honest with a caveat: layouts need different alignment (2D — 256 KB blocks, 4D — 2 MB), so corpus coverage differs — figures are per each layout’s covered share.

Scripts: probe_matrix_scan.py + matrix_scan_dump.py (strides in the header; the cross-device union exchanges hashes only, MTX-scan-crossdevice-v1); cube hash blake2b-96; saturation curve checkpointed every 500 MB; all figures measured, the extrapolation labeled.

Wave 1 — research complete

Wave-1 deep research is complete: each of the 24 hypotheses carries a real measured verdict (full-cost charged accounting, falsification test executed, script+SHA on the card). GO means the mechanism works and is measured; NO-GO means it is closed by measurement or strict arithmetic. Predicted levers remain predictions and are labeled; the measured numbers live in every card below.

NO-GO · 9 GO · 15

shared context (dictionary / fragments) · 4 identity dedup — reference ≪ payload on exact match · 3 structurally cannot win (boundary) · 4 infrastructure cost accounting · 10 near-match + delta · 3

Generated: 2026-08-28T23:41:35Z · db:addressor_hypotheses

Hypotheses · 24

Page 5 / 12

AH-09 GO W1 · 2026-07-13

MPH catalog

If the matrix index is built on minimal perfect hashing (~2-3 bits/key) plus a short fingerprint (8-16 bits) with cryptographic confirmation on hit, then catalog bytes per chunk become small enough for a billion-chunk store to fit in a few GiB, because MPH is near-optimal in memory.

Data class (Z)

catalog infrastructure (meta)

Address target

n/a

Predicted lever

a 10^9-chunk catalog in a few GiB (prediction)

Ceiling category

infrastructure cost accounting

Mechanism

near-optimal MPH memory + a cheap fingerprint prefilter

Falsification test

fp/bytes curve; if a safe fp-rate is unreachable with a 16-bit fingerprint and the catalog eats the amortization -> configuration NO-GO

Full cost (total_cost)

infra line item: catalog_bytes/chunk ~ 2-3 bits MPH + 8-16 bit fingerprint; on hit — confirmation by full hash (network) · amortized: divided across all fleet accesses.

Probe verdict

GO — the compactness claim is confirmed with a REAL serialized structure (no MPH even needed): 2.0 B/key at a 16-bit fingerprint, measured fp 3.96% (matching theory 2664/65536=4.06% — the structure is honest), every positive confirmed by full hash over the network. Extrapolating the measured density to 10^9 keys = 1.86 GiB (labeled arithmetic on measured data). Sensitivity: fp8 is useless (fp~100%), fp12 intermediate (47.8%), fp16 is the operating point. The falsification condition (catalog eats the amortization) did not fire: 2 B/key << the 34.3 B/key wave-1 model.

Measured result (raw research log; descriptors partly in the RU original)

date: 2026-07-15
probe: DR-catalog-bloom-v1
verdict: GO

structure

реальный bucketed sorted-fingerprint индекс (packfile-idx стиль) над 681 973 реальными чанк-хэшами devs∪prod

verdict_note

GO — the compactness claim is confirmed with a REAL serialized structure (no MPH even needed): 2.0 B/key at a 16-bit fingerprint, measured fp 3.96% (matching theory 2664/65536=4.06% — the structure is honest), every positive confirmed by full hash over the network. Extrapolating the measured density to 10^9 keys = 1.86 GiB (labeled arithmetic on measured data). Sensitivity: fp8 is useless (fp~100%), fp12 intermediate (47.8%), fp16 is the operating point. The falsification condition (catalog eats the amortization) did not fire: 2 B/key << the 34.3 B/key wave-1 model.

sensitivity_fp_bits

	bytes_per_key	measured_fp_pct
fp8	1	99.997%
fp12	2	47.809%
fp16	2	3.962%

extrapolated_1e9_keys_GiB_at_fp16: 1.86

AH-10 GO W1 · 2026-07-13

Sender-side Bloom prefilter

If the sender keeps a local Bloom filter of the catalog (fp ~1%), then network lookups on a typical user mix fall to roughly hit_rate x (1+fp) of chunks, because known-new chunks are rejected locally without a round trip.

Data class (Z)

typical user mix

Address target

n/a

Predicted lever

orders of magnitude fewer network lookups (prediction)

Ceiling category

infrastructure cost accounting

Mechanism

local rejection of misses

Falsification test

matrix churn model vs filter-update traffic; updates costlier than the saved lookups -> NO-GO

Full cost (total_cost)

infra line item: client filter ~1-2 bytes per catalog chunk + UPDATE TRAFFIC under matrix churn · savings: lookup round-trips.

Probe verdict

GO — a real filter against a real device stream: 70.16% of network lookups cut (151,189 -> 45,116; 44,828 true hits, 288 false passes), measured fp 0.298% at the 1% target. The falsification condition (filter updates costlier than the savings) was executed with the MEASURED drift (1.3%/month): churn traffic is 0.11% of one full scan's savings — it did not fire, by ~3 orders of magnitude. A 1 MiB filter for the fleet catalog is negligible in total_cost.

Measured result (raw research log; descriptors partly in the RU original)

date: 2026-07-15
probe: DR-catalog-bloom-v1
verdict: GO

structure

реальный Bloom-фильтр над 681 973 ключами; реальный поток устройства arcana-www (151 189 уникальных чанков)

sensitivity

	bits_per_key	filter_bytes	measured_fp_pct	lookup_reduction_pct	churn_over_savings_pct
target_1pct	10	1.00 MiB (1 048 576 B)	0.298%	70.16%	0.11%
target_0.1pct	15	2.00 MiB (2 097 152 B)	0%	70.35%	0.16%

churn_source: ИЗМЕРЕННЫЙ дрейф KB 1.3%/мес (AH-12)

verdict_note

GO — a real filter against a real device stream: 70.16% of network lookups cut (151,189 -> 45,116; 44,828 true hits, 288 false passes), measured fp 0.298% at the 1% target. The falsification condition (filter updates costlier than the savings) was executed with the MEASURED drift (1.3%/month): churn traffic is 0.11% of one full scan's savings — it did not fire, by ~3 orders of magnitude. A 1 MiB filter for the fleet catalog is negligible in total_cost.

Page 5 / 12

Where the Addresser cannot beat local compression

Six boundaries are fixed by wave 1 as explicit anti-hypotheses and no-win zones: unique high-entropy data (personal media, encrypted streams — nothing repeats globally); tiny unique files below the inversion point (fixed catalog costs exceed any saving); mathematically generated matrices (an address into an exhaustive or random pool costs at least as much as the content itself); a byte histogram or hash used as the data carrier (order is lost and buying it back costs the file’s entropy); long-tail content fetched roughly once (the first transfer is never repaid); and fragment schemes whose gain collapses into what a shared dictionary already provides. A tiny fixed-size reference per object (tens of bytes) is honestly achievable only on an exact match against a catalog that already stores the object; everywhere else the win criterion is simply that the reference plus all charged costs stay well below the payload it replaces — the reference size is a metric to minimize, not a hard gate.