igneum-network/igneum
1
0
Fork 0
Code Issues Pull requests Releases Activity
2391 commits 357 branches 1 tag 2 GiB
Commit graph

162 commits

Author SHA1 Message Date
igneum-labs
bb52ac640f Class v6 family gate, the 17:00 cut: the harness as a plain diff at 3dc3117c (the founder-strings check refuses the patch form's author lines) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 11:35:25 +00:00
igneum-labs
a421b896f5 Class v6 family gate, the 17:00 cut: the last-resort rows corrected for class v5's verified scan (the finding at the corner is the correlation of one era's attempts, 1,000x the independent estimate, not an unchecked program); the harness patch series at 236ef3a5 (the lossy-share curve and the scan's verdict column) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 11:27:11 +00:00
igneum-labs
17a9bf682d Class v6 family gate, the 17:00 cut: the measured coverage (24,000 drawn eras through the ring-B tests on build-1 and build-3, 54 core-hours), per stratum and per axis, with the bound arithmetic on the counts; four measured rows (the lossy corner of B = 4 exhausts the 256 cap in 1.2 percent of eras, the band capped at the lossy families; the (c''') floor's per-width calibration; the shape axis's draw cost; the era-stride bias at bit R in half the eras of every stratum); the census rows, logs, scripts and the harness patch series under logs/ 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 11:25:26 +00:00
igneum-labs
8551379fc2 Merge build/master 0d7412ac into class-v6-family-gate (export-exclude.txt: both research paths kept) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 10:32:44 +00:00
igneum-labs
eb71eaf399 Merge class-v6-history 2b76937c into master (gate: green on 2b76937c, recorded by tools/ci/pre-push.sh; landed on the build mirror) 2026-10-08 10:26:13 +00:00
igneum-labs
1dcb3882e1 Class v6 family gate (lane D), first cut: the parameter table with bands and margins, the sampling bound (rule of three, the union bound over the per-era tests' measured miss rates, the stratified corners), the three rings of per-era tests with known-failed shapes and clean-seed costs, the board over the testnet period and the proof-of-testing record format, the honest line; the harness running on the boxes and its first 16-era reading (the era-stride class at 130 to 511 sigma on one address bit in 7 of 16 drawn eras); docs/analysis/class-v6 excluded from the public export 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 10:25:17 +00:00
igneum-labs
2b76937caf Class v6 history lane, first cut: every ASIC-resistant proof of work and its chips by mechanism, the four-layer mapping with a number per chip class, and the three lessons that bind on v6 
docs/analysis/class-v6/history.md: 21 chip rows by mechanism (what each chip specialised, what the design missed, the timeline, the v6 layer, closed or not, the per-joule number); in-depth sections for Ethash (the E3's DDR3, Linzhi's 4.4 GB beside 64 compute units, the Jasminer X4's DRAM hybrid-bonded onto a 40/45 nm die, the E9), ProgPoW (the Least Authority and Rao audits read from the PDFs, Kik, Linzhi's claim, the adopters), RandomX (the X5's RISC-V board at 1.46x, the X9 announced December 2025 and withdrawn May 2026 with zero units, RandomX v2 released 25 March 2026 with activation pending, the Qubic episode as the detector's lesson), Cuckoo (the GRN1's 512 MiB SRAM die, the G32, the iPollo G1, the 2025 bounty), Equihash (the Z9's single die, the Fudan parameter-following design), Scrypt and Argon2 (the BM1485's on-die scratchpads, Percival's lookup gap, MTP), the no-chip hashes, kHeavyHash, CryptoNight, X16R, Lyra2REv2, the compute rows. The four layers against the history layer by layer, with the E3's death dates, Autolykos's growth rule, the design-cycle and FPGA-compile clocks, and the honest verdict that the stored-dataset chip (class C) is not closed by any layer: 5.1x per joule at zero premium, 2.1x with the shadow at k = 1. Two corrections to the 5 October file (the X9, RandomX v2) and one claim marked unverified (the 5x fork-surviving chip). First cut for 15:00 UK; the full report by 09:00 UK on 9 October.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 10:19:10 +00:00
igneum-labs
93e55f9d43 Class v6 research lane B: the hardware future, five years out (first cut and the full report in one file) 
docs/analysis/class-v6/hardware-future.md: one table over PIM and PNM (HBM-PIM, AiM, LPDDR5X-PIM, UPMEM), HBM3E, HBM4 and the custom HBM4E base die, fine-grained DRAM on logic, LPDDR6, 3D DRAM, RLDRAM 3 and Folded Banks, FPGA with HBM2e, wafer-scale, CXL, optical I/O, chiplets and UCIe, and SRAM at N2: energy per dependent read, latency, cost per GB, availability to a non-hyperscaler, the k_read and shadow k bands, the edge per joule at zero shadow against the 5090 and the M5 Max, and the one layer of the four that blunts each with a number. The three findings: a 2 GiB SRAM full store on one N2 reticle is the strongest five-year chip (13x to 17x per joule, about USD 0.3 per MH/s) and no layer reaches it (layer 2 moves its capex, not its joules; the shadow at k = 0.5 holds 4.8x); per-bank PIM is blind (1.6 percent of reads in-bank at 2 GiB) and the custom HBM4E base die is the f = 1 chip with its controller in the stack (6.5x to 14x), untouched by layers 1 to 4; the denominator is the wrong card (the M5 Max at 0.78 microjoules halves to thirds every chip edge). Section 5 carries the k bands for the research lane's chip rows; section 7 the four decisions with defaults and deadlines; section 9 the sources with URLs and dates. tools/ci/export-exclude.txt: docs/analysis/class-v6 is research, not public export.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 10:17:50 +00:00
igneum-labs
29239a2296 attack-pass: the status board, F8 row, gate line and re-gate paragraph read the tail as attributed 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 08:49:21 +00:00
igneum-labs
7ac25503ed attack-pass: AP-F8-1 tail attributed (p4, p8, p10, p34 as per-site bucket concentration at a narrow-window site; the largest-bucket bound named for a next class) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 08:48:03 +00:00
igneum-labs
b7ea50ef69 Chip model 5.11 and new-pow 5.1: the per-MAC figures corrected by the per-lane tile count (32 MACs per tile, not 1,024; the 4090's 0.056 pJ is 1.8 pJ, the 5090 reads 2.9 and 1.5 at the shadow's premium), the tensor-tile column withdrawn (a chip's k on tile work 0.03 to 0.3, below the ALU shadow's); Counter ASIC 3.0 status: the microbench's correction, the night's closing sentence on measured rows, the CA4 file's commits 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 06:08:07 +00:00
igneum-labs
15e70626fb attack-pass lane (d): F1 shadow redundancy on the frozen class v5 tip dba17a0a, 10^5 programs, max 4.688 percent, 0 mismatches, PASS (the 0.3.24 gate line; AP-F1-1 FIXED-AND-PASSED on v5) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 04:22:29 +00:00
igneum-labs
7c1784e64c attack-pass: F1 record section 12, the census flush and its known-failed test (2,000 rows after a kill at 2,000, PASS) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 04:08:52 +00:00
igneum-labs
fcf92fb806 attack-pass: founder-strings scrub on the pass record (the gate's third never-push class) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 03:10:09 +00:00
igneum-labs
0f6419ff31 Merge remote-tracking branch 'build/master' into attack-pass 
# Conflicts:
#	docs/analysis/attack-pass-2026-10.md
 2026-10-08 03:08:58 +00:00
igneum-labs
9a47418517 attack-pass lane (d): F1 on dba17a0a running at the 03:06 UTC reading (15 cores, projected end about 09:00 UTC); the census harness's no-progress gap named, the progress line and flushed partial owed 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 03:07:23 +00:00
igneum-labs
d626b6b083 attack-pass lane (d): F9 exhaustion count on the frozen class v5 tip dba17a0a, 10^5 seeds, 0 exhausted, PASS (the 0.3.24 gate line) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 02:38:51 +00:00
igneum-labs
ab71510c18 attack-pass F2 record: in-house wording 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 01:59:00 +00:00
igneum-labs
12f2688039 attack-pass: in-house wording (no outside reviewers named or implied); the open tail is named in the public report 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 01:58:34 +00:00
igneum-labs
397bee40e0 Counter ASIC 3.0 status: the lanes' own 00:00 lines, the spec-text re-derivation ordered, the class v5 packs test's derivation-text diff and its disposition, the one 0.3.24 kit named (e6c088bb); chip model section 5: the second partial-store correction from adv-cache-2's window-layer reading 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 00:11:21 +00:00
igneum-labs
d2f49114c8 Counter ASIC 3.0 status: the fast-time PASS on the re-cut c9e385eb and the move's interim gate, the 61-row era reading and the class v5 attempts census, the adv-accept 00:00 BST reading, the per-load class closed; chip model 5.11's 200 M clause marked as resting on the closed construction 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 22:59:04 +00:00
igneum-labs
05fcbb84c4 attack-pass lane (d): F4 on 19c60cf9 PASS, AP-F4-1 FIXED-AND-PASSED against class v5 (0 of 2^24 over 1.1x) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 22:33:49 +00:00
igneum-labs
3231a80fe2 Site: the pages rebuilt on the record branch (the regrouped bench table and the outside-check text as built from this tree) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 22:16:30 +00:00
igneum-labs
c779d0308e attack-pass lane (d): F8 on the frozen class v5 tip dba17a0a PASS (61 of 64, the three residue seeds, worst 1.50x); F9, F1, F4 rows queued 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 22:06:35 +00:00
igneum-labs
25b1a4e149 attack-pass F4: reconciled with adv-mixer-2 (the carry digit at position 32 counted here and not there; agreed median 226, 15 days a century, worst 2050-04-28 at 1.113x; the rule takes adv-mixer-2's form) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 20:39:42 +00:00
igneum-labs
5fd9599714 Chip model 5.10: class v5 on with the shadow at zero leaves the stored-dataset chip at 5.1x (GDDR7) to 9.1x (HBM3) per joule because the node is a farm cost, not a chip cost; the shadow stays the only lever in the model (the founder's question, 7 October 2026) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 20:29:26 +00:00
igneum-labs
49d6d52a76 Chip model section 5: the pebbling-optimum correction from the in-house pass (9.3x at f = 1/64, not 17.4x; the full-store verdict stands); Counter ASIC 3.0 status: the (c''') test green, the pool yield, the restart-step IBD deadlock and its fix, the pass's readings and findings, the pod ledger, the fleet table 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 19:37:23 +00:00
igneum-labs
b65d63dacd attack-pass lane (d): F8's pre-freeze v5 reading (equal to sub-version 3 seed for seed), the yield, and the frozen-tip gate plan 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 19:36:09 +00:00
igneum-labs
cc3a5b334f Merge scrub f3be8d2e into master (gate: green on f3be8d2e, recorded by tools/ci/pre-push.sh; landed on the box mirror under the exception declared by main: main's ruling, 7 Oct 2026 19:5x UK: the GitHub account is suspended, lanes land on the box mirror's master, the box gate stamp is the verdict; GitHub gets the fast-forward when it answers) 2026-10-07 19:28:40 +00:00
igneum-labs
225ed25303 attack-pass lane (d): the hand-started class v5 runs killed on the coordinator's rule, re-queued through lease pool; F8's first 11 seeds equal sub-version 3's 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 19:21:50 +00:00
igneum-labs
6807646818 Merge site-22 f499c291 into master (gate: green on f499c291, recorded by tools/ci/pre-push.sh; the full gate runs in CI on this merge) 2026-10-07 19:17:32 +00:00
igneum-labs
f3be8d2ead Merge remote-tracking branch 'box/master' into scrub 
# Conflicts:
#	CLAUDE.md
#	docs/plans/counter-asic-3-status.md
#	docs/plans/release-0.3.21.md
#	docs/plans/release-0.3.22.md
#	tools/ci/merge-to-master.sh
#	tools/ci/pre-push.sh
#	tools/ci/red-watch.mjs
 2026-10-07 19:16:14 +00:00
igneum-labs
f499c2919b Site, 7 October 2026 evening (the project lead's four asks through main, one landing): (1) the bar's nav to the right: the mark and the devnet pill on the left, Mine, Network, Learn and the controls right-aligned with the same gaps and sizes, each open panel anchored under its button (the script sets --panel-left) and a panel that would run past the container aligned to its button's right edge, the phone sheet unchanged; (2) no lateral scroll: tools/ci/scroll-width-check.mjs on the gate (scrollWidth equals clientWidth on every route at 390, 768, 1024, 1280 and 1440 in both themes; only a table, code or diagram scrolls inside its own overflow-x box; self-test with a 100vw plus 20 px fixture known-failed first); (3) LG-5, the launch text: the front page leads with the three wants (hardware that stays useful, income without a cliff, a fair supply) and the live network with finality is the second screen; the litepaper's regulatory note reads "Not legal advice." alone over its eight sentences from future.md 8.3 (the project lead: counsel not yet; the string "counsel is engaged" is in both forbidden-strings lists and gone from the research heading), the pay sentence stays; the journey's phase line reads Devnet 3 live, testnet next; (4) LG-1, /income: docs/analysis/income-tiers.md published as a page by tools/launch/income-page.mjs (checked on the gate against the document), listed under Learn, labels on every number, no price of IGN anywhere 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 19:12:04 +00:00
igneum-labs
a43355dca3 Merge attack-pass: the internal attack pass F1 to F10 before the cryptanalysis freeze (every row PASS or FIXED-AND-PASSED; the ten records and harnesses; AP-F1-1, AP-F4-1, AP-F5-1, AP-F8-1/2/3 and the two operating hazards) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

# Conflicts:
#	docs/bench-log.md
#	docs/evidence.md
 2026-10-07 18:54:22 +00:00
igneum-labs
9dd834751e attack-pass lane (d): class v5 re-runs, F4 PASS (byte-identical to v4); F8, F9, F1 running 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 18:51:55 +00:00
igneum-labs
8bb638f843 Pre-public scrub, the text pass (7 October 2026, 19:5x UK): no founder name, personal login, earlier business or personal address in any tracked text file, and a gate check that keeps it so 
The sweep (main's item 1): 199 tracked text files, 783 lines. The founder's full name, first name and possessive become "the founder" (sentence starts capitalised); the lowercase operating-system user name in WSL paths and commands becomes <user>; the second owner login becomes "the second owner login"; the three earlier businesses and the two other brands become "the other business", "the earlier entity", "the earlier business" and "another brand"; the Chrome profile rule names the igneum.network profile, not the profile's label. The standing commit login igneum-labs is not a founder term here: the fresh-repository step renames it in the history (docs/plans/history-rewrite.md, tools/repo/fresh-repo.sh).

The patterns never appear in plain text in the tree (a plaintext list would be the hit): tools/ci/founder-strings.b64 (perl regex, tab, a sample per row) is read by tools/ci/founder-strings-check.sh (every tracked text file, perl, known-failed first: the self-test plants each row's sample in a fixture and the hit must name the file), by tools/community/discord-hooks.mjs (the guard's founder and business rows; the test takes its fixtures from the samples) and by tools/repo/fresh-repo.sh (the business names of the rewrite rules). site/forbidden-strings.txt carries the same patterns as b64: lines, decoded case-insensitive by site/scrub.mjs and tools/ci/launch-gates-check.mjs (whose fixture now plants an encoded made-up name). The check runs in the gate's tree checks on every merge.

Not in this commit, by main's word: the 105 commit messages and 40 personal-identity commits that need the history rewrite (listed, not run), and the secrets found by gitleaks over the history (reported with owners).

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 18:39:50 +00:00
igneum-labs
e80978adc1 Merge origin/master 0caa3cc5 into ca3-v4-amend (igneum-pow kept byte-identical to the frozen c79f6d96 per main's stop-the-line ruling; the ledger keeps both sides) 2026-10-07 16:40:29 +00:00
igneum-labs
ade7cf6edc attack-pass: F8 FIXED-AND-PASSED in class v4 sub-version 3 (c79f6d96: 60 of 64 under 1.2x, the named four-seed tail, 0 exhausted); the pass record reads PASS or FIXED-AND-PASSED on every row 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 16:03:18 +00:00
igneum-labs
a1e45767f0 attack-pass F7: sub-row (a) PASS, the era VDF in the node and the re-roll harness silent against it (0 of 6 cuts); the freeze precondition met 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 15:33:17 +00:00
igneum-labs
6433f3ab95 Era VDF lane (7 October 2026): spec 04 sections 4.2, 4.4, 4.5 and 4.6 completed for the era path (the scheme byte and the hash-chain fallback, the cut rule under the O-4.3 reading, the day-of-blues input, the delay, the seed, the era constants as measured), spec 01 section 1.13.1 and era-layout.md section 8 updated, open items O-4.8 closed and O-4.10 to O-4.12 added, FUD ledger F11 updated, two decisions for the project lead in ledger-decisions.md, the record docs/analysis/era-vdf-2026-10-07.md (what was built, the parameters and measured rates, the harness gate with both runs, the cut rule for the finality lane, the verify gate, the tiers), the F7 re-roll harness against the real era cut (tools/era-vdf/reroll.mjs: --vdf off fires 6 of 6, --vdf on silent 0 of 6; the adversary evaluates asynchronously with the node's own code and is found by the node's own log line), the fast-time file's era fields 2026-10-07 15:24:22 +00:00
igneum-labs
03f32f1cd2 attack-pass F8: sub-version 3's two commits, the (c'') threshold evidence and the c79f6d96 census line in the record 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 15:15:44 +00:00
igneum-labs
629ec41c5d attack-pass: AP-H2, a rebuild on the shared binary path reached two runs in progress; the launcher now runs from a copy in the run's scratch dir 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 15:00:26 +00:00
igneum-labs
b743a1f5da attack-pass F8: sub-version 2 re-gate verdict FAIL (9 of 64, worst 4.82x), the per-seed table and the clean per-site spread; sub-version 3 is the target 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 14:22:42 +00:00
igneum-labs
7c45f3a21c attack-pass F8: p23's residual localised and explained (or-then-xor same-operand AND mask over a mulhi output, lineage-blind); the distinct-index ratio rule for sub-version 3 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 14:20:14 +00:00
igneum-labs
2a8842bee5 attack-pass: AP-F8-3, the acceptance rule never executed the shadow block (the root of the residual classes), found by the hash lane; fixed in sub-version 3 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 14:07:25 +00:00
igneum-labs
d290834da3 attack-pass: AP-F8-2 exhaustion half FIXED-AND-PASSED at 773100c4 (0 exhausted in 10^6, max attempt 35, 4 past 31) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 14:05:02 +00:00
igneum-labs
a75804435a attack-pass F6: PASS, the worst of 10^5 class v4 programs reads 8.708 ms on the half-core proxy (1.29 ms under the gate), batches B and C under the per-core lease 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 14:00:55 +00:00
igneum-labs
e903bc32c7 attack-pass F8: AP-F8-2 closed at 773100c4 (0 exhausted in 650k, max attempt 35); sub-version 2 fails the hot-set gate on lineage-fresh constants (8 of 39) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 13:57:38 +00:00
igneum-labs
a2d1da7527 attack-pass F2: k = 3 and 4 lines closed at the solver cap (no trail under 29 to 35 and 39 to 47 differential, 24 to 28 linear); row PASS effort-bounded 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 13:36:58 +00:00
igneum-labs
51bf972956 attack-pass: AP-F8-2, a chain-shaped epoch seed exhausts 32 attempts under sub-version 2's rule (a'), a liveness class 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 13:33:40 +00:00