igneum-network/igneum
1
0
Fork 0
Code Issues Pull requests Releases Activity
2450 commits 395 branches 1 tag 2.7 GiB
Commit graph

21 commits

Author SHA1 Message Date
igneum-labs
6d0a52ea03 Litepaper and evidence: the hash's reads stated as measured (8 October 2026, lane D's family harness at 2^20 over 4,900 drawn eras and adv-cache-2's reading): "dependent reads spread over a multi-gigabyte dataset" in the lead, the measured clause after the chain-of-random-reads sentence (the 0.995 floor on every accepted program; about half of epochs with one biased address bit at the era's stride rotation; about 1.6 percent of a hash's reads to a chip storing half the dataset, nothing to a full store; the fold in the next class), evidence row 18 with its sources; ledger AP-F8-7 (Open, priced), the public ledger and the ledger page regenerated. Bench table: class v4 watts rows 7 to 10 (RTX 3070 196.4 W under a 193 W cap, +28 percent; RTX 3070 Ti 267.6 W uncapped, +48 percent; RTX 4060 no watts on any rented host; the 3080 Ti's deep-capped pair in its note, the first cap that costs rate, the uncapped rate standing) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 10:48:11 +00:00
igneum-labs
ba08d6b8a3 Bench table: the class v4 watts rerun, first six rows (RTX 5070 176.1 W, 4070 Ti Super 137.8, 4070 Ti 155.3, 4070 Super 108.4 under a 110 W cap, 4060 Ti 16 GB 102.3, 4080 185.6; each with the class v3 watts paired on the same host as the shadow premium, +9 to +56 percent for no rate; the 5070 and 4070 Ti Super rates as ranges by board), MH per watt ranked again on those rows 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 10:39:24 +00:00
igneum-labs
eca96c4799 Bench table: the fleet's watts are class v3 (the sweep lane's read of its pack, 8 October 2026: program_class v3, generator 3): every model-sweep row and every 7 October fleet row carries its watts as "(class v3)", its MH per watt hollow and unranked, the per-watt lede claims drawn from class v4 rows only, the sweep line stating the fact and the measured gap (a 5070 Ti at 141 W on v3 and 224 W on v4 at one rate); the rates stand (memory-bound, the same under both) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 10:22:49 +00:00
igneum-labs
443ac757fe Bench table: the model sweep's last rows (RTX 4070 Ti 31.24 MH/s at 95.4 W; L40S 56.41 at 220.7 W in the datacentre table; RTX 4080 Super 42.60 at 134.9 W); the GTX 1080 Ti row's flag cleared by its second host (five of five self-tests pass, the failure was the first host's VRAM); the page carries the sweep's two findings as one line (class v4 cap-proof; class v5 level on uncapped cards and 17 to 18 percent under on two power-capped hosts, so a v5 rate needs the power state beside it) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 10:05:26 +00:00
igneum-labs
f97ed89621 Bench table: the model sweep's rows 16 to 19 (RTX A6000 53.64 MH/s at 239.1 W in the datacentre table, v5 level with v4 on an uncapped sm_86; GTX 1080 Ti 16.26 at 124.3 W with the one-in-three cache self-test flag until a second host decides; RTX 4060 19.09 with no watts, the host has no power sensor; RTX 2060 14.13 at 120.6 W, the floor of the list; rented and untuned, 8 October 2026) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 09:57:36 +00:00
igneum-labs
bd92081011 Bench table: the model sweep's rows 11 to 15 (RTX 2080 Ti 26.01 MH/s at 224.6 W and RTX 2070 Super 19.09 at 166.0 W, the Turing per-watt floor; A100 SXM4 138.10 at 268.6 W and H100 SXM 248.30 at 411.7 W in the datacentre table; RTX A4000 31.41 at 112.8 W under a 114 W host cap, its class v5 reading held off the page until an uncapped card confirms it; rented and untuned, 8 October 2026) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 09:47:14 +00:00
igneum-labs
97176c3554 Bench table: the model sweep's rows 5 to 10 (RTX 3070 37.09 MH/s at 142.3 W; RTX 3070 Ti 38.98 at 177.1 W; RTX 4060 Ti 16 GB 20.07 at 79.2 W; RTX 5070 51.81 at 110.8 W; RTX 4070 Super 31.24 at 100.2 W under a 110 W host cap; RTX 3060 Ti 31.70 at 100.4 W; rented and untuned, 8 October 2026, the class v5 kit's fingerprint matched on each) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 09:39:57 +00:00
igneum-labs
af528096e9 Bench table: the model sweep's rows 2 to 4 (RTX 4070 Ti Super 41.24 MH/s at 142.6 W; RTX 3080 Ti 58.86 at 293.5 W; RTX 4080 40.71 at 128.6 W; rented and untuned, 8 October 2026, the class v5 kit's fingerprint matched on each, the hash at 0.99 to 1.00 of each card's read ceiling) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 09:32:03 +00:00
igneum-labs
208860af9f Bench table: the model sweep's RTX 5070 Ti row (78.45 MH/s at 135.9 W, 0.577 MH per wall watt, rented and untuned, 8 October 2026; the class v5 kit at the same rate, fingerprint matched; the hash at 0.99 of the card's read ceiling); the /miners lede recomputes from the data as three separate claims (best desktop card by rate, best stock MH per wall watt, best tuned), so a rented stock row never reads as tuned 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 09:25:32 +00:00
igneum-labs
e1401ffd66 Bench table: the Intel Arc B580 row stands, with the 8 October re-read in its note (the installed 0.3.21 worker passes its self-test on this card, 96 of 96 lanes, 54 re-checked blocks at 10.58 MH/s; the kit build that failed lacked the Intel rotate-fold rewrite, a build fault and not an Arc result); /miners rebuilt 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 08:46:43 +00:00
igneum-labs
3c152ae10c Bench table: the RX 9070 XT row takes the app's own power reading at the stock point (202 W at 18.9 MH/s, 0.093 MH/W; the G1 ladder's 18.8 to 19.2 as the rate; the status row of 8 October 2026), tune state "no lever" (no AMD knob in 0.3.20), the class v4 cost unchanged, Hive values none; /miners rebuilt 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 06:28:55 +00:00
igneum-labs
1bc1752d90 Bench table: the RTX 5080 row states the team's clock-lock pass as the card's figure (71.4 MH/s stock, 71.2 tuned at 146.6 W at a 1,100 MHz core lock, +83 W unlocked and +41 W at the best points for class v4, Hive core 1100) with the rented-fleet sampler reading and its 110 W gap kept in the note as the open question; /miners rebuilt 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-08 02:13:09 +00:00
igneum-labs
9cfa334894 Site: the GPU bench table presents the same data the way a buyer reads it (the founder, 7 October 2026): cards you can buy first, sorted by hash rate; datacentre cards collapsed; the Apple row's chip-watt basis marked and unranked under MH per wall watt; one 5090 row with the class v4 knee as its class v4 figure and the efficiency-pass rows in its note; the best desktop card line above the table 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
(cherry picked from commit 2336a3c5cb2cb9a652c494ab6f9334319f8752f3)
 2026-10-07 22:16:30 +00:00
igneum-labs
64477ff098 Site: the bench table's v4 cost and tuned columns are one clause each (the sentences live in the detail row), so the five-number row fits the article box (the 22:10 BST capture read the cost column wrapping to six lines) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 21:15:14 +00:00
igneum-labs
dac2e7b50e Counter ASIC 3.0 status: the merge's duplicated hunk resolved; the bench table's efficiency-pass rows name the desk machine, not its number (the site scrub rule) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 19:50:43 +00:00
igneum-labs
cd4a62c6ef Site: the 5090's locked point (1,400 MHz, 134.98 MH/s at 316.3 W, 0.427 MH/W) and the locked class v4 premium beside the unlocked one in the bench table; Counter ASIC 3.0 status: the efficiency pass's full table, Devnet 3's first lock, the fleet table, the lease-pool rule, the 0.3.24 clock 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 19:22:03 +00:00
igneum-labs
bc09ad66b9 Site: the GPU bench table carries the Hive flight-sheet values at each measured tune point (the 5090 1,854 MHz, 13,801 MHz, 460 W; the 4070 1,863 MHz, 10,251 MHz, 100 W; stock elsewhere) and the two-line Hive note (main, 7 October 2026) 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 18:49:27 +00:00
igneum-labs
67d7da5510 Site: the GPU bench table carries the fleet's rows (every rentable card on the hash, 7 October 2026: 5080, 3070 and 3070 Ti, 3080 and 3080 Ti, 3090 and 3090 Ti, 4090, 3060 and 3060 Ti, 4060 Ti, 4070 Ti, 5060, 5070 and 5070 Ti, A5000, L40S, A100 PCIe and SXM, H100 and H200, B200, RTX PRO 6000) with watts, MH per watt, driver and the tuned state 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 18:04:04 +00:00
igneum-labs
9282712ffc Site: the GPU bench table carries every measured card (5090 stock and tuned, 4070, 9070 XT, M5 Max, 5060 Ti, Arc B580, H100 SXM) with watts, MH per watt, the class v4 cost per card and the tuned state; the fleet's 3070, 3080, 4090 and 5080 rows follow 
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 17:57:47 +00:00
igneum-labs
052c935b7b RTX 5060 Ti (16 GB, Razer Core X V2 Thunderbolt) on PC 2: detection, G1 bit-exact on class v4 sub-version 1 and the v3 control, the 10-minute stock window, the bench-log entry, the card-picker entry and the public table row 
The first 16 GB card and the first Thunderbolt card. The app's first poll listed it (vendor nvidia, 16,311 MiB,
driver 610.47 shared with the 5090, kind discrete: the app does not know it is an eGPU) and started a miner on
it by itself. Job run-5060ti-bench-20261007-b (tools/bench-5060ti/pc2-5060ti-bench.ps1, the card alone under
the runner's --cards-off, the 5090 mining, no power limit written: 180 W default read back unchanged, PC 2 lost
power twice that day): G1 PASS with 867dbc45cfb36b4d (sub-version 1) and 90f794dd556f7a3b equal to the Mac's;
the window 1,105 dispatches of 2^24 in 602 s at 30.882 MH/s, 114.8 W mean at utilisation 90 percent and over,
0.269 MH/W, PCIe 4.0 x4 through the enclosure (the hash is latency-bound, the link costs nothing measurable).
Owed: the efficient point (the app's Ember Tune; PC 2's Power Helper refuses every request since the restart)
and the prove-beside read (the shipped WSL2 host has no IGNEUM_CUDA_DEVICE selector; the floor server is on
PC 2; the prover-floor host is the cut). The kit (tools/bench-5060ti/make-kit.sh) takes the sub-version 1 packs
from ca3-v4-amend aa31c396, since master's packs-ca3-v4 is the 6 October stream.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-07 15:05:47 +00:00
igneum-labs
12a4dae0f5 App engine: watchdog per card and per node, WORKER FAULT and mismatched= read from the miner; public GPU bench table and the reliability test harness 
Review round 4 X21 and the project lead's levers 5 and 6 (4 October 2026 evening):
- src/watchdog.rs: CardWatch (no status line for 90 s, or hash rate 0 for 60 s while the node is synced: one
  restart, then the card is marked faulted with the reason in the UI and the log; the miner's own worker restarts
  suspend both rules until ready, so there are no double restarts; exit 43 counts as a watchdog restart) and
  NodeWatch (our node silent for 120 s is restarted in-process through the restart kind the remote jobs use, with a
  growing delay). State machines with no clock; 11 unit tests replay recorded STATUS and WORKER FAULT lines.
- engine.rs reads STATUS mismatched=, faults= and the WORKER FAULT lines onto the card (faults, mismatched, message);
  a faulted card is not restarted until the user changes its settings or resumes mining; other cards keep mining.
- site/miners.html (build.mjs, scrubbed like /bench, rows from site/miner-bench.json): card, generator version, best
  MH/s, MH per watt where measured, miner version, date, source, measured by the team or reported by the fleet. In the
  navigation beside the engineering log on every page and in the sitemap. One line says there is no other miner to
  compare with.
- tools/reliability: fake-worker.mjs (the serve protocol, misbehaving on command), run.mjs (miner guards on a private
  test network, ports 29950+) and app-run.mjs (the app watchdog end to end, ports 29960+).

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-10-04 19:20:49 +00:00