diff --git a/docs/evidence.md b/docs/evidence.md index 431d7283d..33ed89071 100644 --- a/docs/evidence.md +++ b/docs/evidence.md @@ -36,7 +36,7 @@ Versions in the table: "the release manifest" is `site/release-manifest.json`, s | 3 | The 64-register window per lane costs a GPU under 1 percent of rate at stock, and at most 5 percent per load with the liveness chain | The litepaper (class v6); `docs/plans/igneum-2.0.md` D1 (the placed 64-register rows) | TEAM-REPORTED; tested by the team | `docs/analysis/class-v6/connected-state.md` section 4; `docs/design/class-v6-rotating-family.md` section 10.0e | The class v5 nvcc harness and the kit worker, both packs on the same card minutes apart, 250 batches of 2^24, nvidia-smi at 1 Hz, vectors PASS on every row (`connected-state.md` section 4); the per-load rows of the full chain against the base (`class-v6-rotating-family.md` 10.0e) | 8 October 2026, rented RTX 5090 (575 W cap) and RTX 4090 (450 W cap) at stock: energy per hash +0.6 percent on the RTX 5090 and -0.9 percent on the RTX 4090, inside the run-to-run noise; under 1 percent of rate; 80 to 87 registers per thread, no spill (all measured). Per load: RTX 5090 16.7 nJ base, 17.6 nJ full chain; RTX 4090 26.0 nJ, 27.0 nJ (measured). The lock row on the project's own rigs is owed | none yet | | 4 | Reorganising the same work around live state (the connected-state variant, experiment D2(a)) does not reduce a specialised chip's edge: KILL as a class | `docs/plans/igneum-2.0.md` D2(a); this page | TEAM-REPORTED; tested by the team (a published failure) | `docs/analysis/class-v6/connected-state.md` (the verdict, section 6) | The census, liveness and GPU rows in `connected-state.md` sections 2 to 4; the chip side priced on the drawn program by synthesis (a model, never a lower bound) | 8 October 2026, verdict 17:25 UK: the window is necessary (63 of 64 registers live at every address, measured) but only its width reaches the chip, +1.2 pJ per lane-op at N5 (synthesised); the window moves the chip's edge 1.10x node for node against a 1.25x gate (modelled); the GPU side +0.6 percent energy per hash on the RTX 5090, -0.9 percent on the RTX 4090 at stock (measured). Rearranging the dependency graph of the same operations moves neither side | none yet | | 5 | "A GPU-secured network for Ethereum-compatible applications and verifiable computation." served on every page | Every page | PROPOSED; designed (served) | `docs/plans/igneum-2.0.md` (the objective: the positioning line) | `node tools/ci/ledger-text-check.mjs`: the sentence pinned (R0) on the home page, the litepaper and this page | 8 October 2026: on this page; the home page and the litepaper carry it as their 2.0 text lands (designed) | none yet | -| 6 | The chip claim as served: "Igneum remains competitive on accessible commodity GPUs even when specialised mining hardware is assumed to exist, remain compatible and seek profit; its security does not rely on identifying that hardware or retiring it through emergency changes." Under it the three statements, separate: energy ("Current modelling estimates a 1.5x to 3.1x energy-efficiency advantage for the specialised designs assessed as complete machines against the GPU tier, from a board on commodity DRAM at 1.5x to an SRAM-store die at 3.1x (1.5x to 2.3x on the GPU's own node)." Per machine on the same node and a node ahead: the DRAM board 1.5x and 1.8x, the hybrid 1.9x and 2.4x, the die 2.3x and 3.1x; 3x to 6x per dollar of hardware at list price. MODELLED: the GPU side measured, the chip's core placed and routed, the rest of the machine modelled, no chip measured, hardware cost approximate within 2x), economic and response capability, rotation an optional improvement. Class v5 derives the dataset from chain state; whether that excludes a specialised design is under evaluation (Deliverable 3), since a design that tracks state is not excluded by staleness. The energy ratio is not the pass criterion: the coexistence model ([docs/analysis/class-v6/coexistence-model.md](https://git.igneum.network/igneum-network/igneum/src/branch/master/docs/analysis/class-v6/coexistence-model.md)) is, and its first run's result is served with its conditions: A specialised supplier may earn a normal return; ordinary GPUs remain sufficiently close in total cost, widely obtainable and useful outside mining that new operators can still compete. On today's modelled rows that holds for the chip anyone can build, a stored-dataset board on commodity DRAM, at a productive life of one to three years: its cost per accepted unit of work sits within the range of the best GPU owner and entrant, it passes six of the seven conditions (a third of the network in boards now costs less than a year's revenue at the final hardware price), and a fleet of it holds a minority of the network with GPU entrants still setting the price. For the SRAM-store die the outcome turns on its hardware cost per unit of work, which the two models price five times apart: at the dearer figure it behaves like the hybrid board and coexists only in a growing network, at the cheaper one a modest fleet takes most of the network once the die exists with its development paid; at either figure it fails the fleet-cost and margin conditions in a flat or shrinking network, and its energy advantage never decides it. A third design, a DRAM board with the hottest half of the dataset in on-board SRAM, sits between the two: it coexists only in a growing network at cheap GPU electricity and fails the cost conditions in a flat or shrinking one, and the dataset's size floor is a real lever on it where it was none on the SRAM die. What holds the die is the investment decision, not the hash; every chip figure here is modelled, not measured, and the chip's hardware cost per unit of work is approximate within 2x. | The litepaper (the chip model) | MODELLED; designed (the bracket modelled; the GPU side tested by the team) | `docs/design/class-v6-rotating-family.md` section 10 (10.0h to 10.0n, 8 October 2026); `docs/plans/igneum-2.0.md` D3 and D4 | the scoring rule in the close (the minimum over workloads of the maximum over free adversarial designs of the GPU's joules per hash over the adversary's, under the 10 percent GPU-cost budget at the lock, the verifier limit, cross-vendor correctness and hardware accessibility); the placed rows are D3's | the GPU side measured: the RTX 5080 at its 1,100 MHz lock 2.06 microjoules per hash and the RTX 5090 at its 1,300 MHz lock 2.33 (8 October 2026, the project's own rigs and rented pods); the chip side synthesised and claimed, its placed gated row pending | none yet | +| 6 | The chip claim as served: "Igneum remains competitive on accessible commodity GPUs even when specialised mining hardware is assumed to exist, remain compatible and seek profit; its security does not rely on identifying that hardware or retiring it through emergency changes." Under it the three statements, separate: energy ("Current modelling estimates a 1.5x to 3.1x energy-efficiency advantage for the specialised designs assessed as complete machines against the GPU tier, from a board on commodity DRAM at 1.5x to an SRAM-store die at 3.1x (1.5x to 2.3x on the GPU's own node)." Per machine on the same node and a node ahead: the DRAM board 1.5x and 1.8x, the hybrid 1.9x and 2.4x, the die 2.3x and 3.1x; 3x to 6x per dollar of hardware at list price. MODELLED: the GPU side measured, the chip's core placed and routed, the rest of the machine modelled, no chip measured, hardware cost approximate within 2x), economic and response capability, rotation an optional improvement. Class v5 derives the dataset from chain state; whether that excludes a specialised design is under evaluation (Deliverable 3), since a design that tracks state is not excluded by staleness. The energy ratio is not the pass criterion: the coexistence model ([docs/analysis/class-v6/coexistence-model.md](https://git.igneum.network/igneum-network/igneum/src/branch/master/docs/analysis/class-v6/coexistence-model.md)) is, and its first run's result is served with its conditions: A specialised supplier may earn a normal return; ordinary GPUs remain sufficiently close in total cost, widely obtainable and useful outside mining that new operators can still compete. On today's modelled rows that holds for the chip anyone can build, a stored-dataset board on commodity DRAM, at a productive life of one to three years: its cost per accepted unit of work sits within the range of the best GPU owner and entrant, it passes six of the seven conditions (a third of the network in boards now costs less than a year's revenue at the final hardware price), and a fleet of it holds a minority of the network with GPU entrants still setting the price. For the SRAM-store die the outcome turns on its hardware cost per unit of work, not its energy advantage: at the reconciled machine cost, set by the power train and the shadow core rather than the die, it fails the cost, hardware, fleet and margin conditions at every point of the band, a modest fleet holds about two fifths of a growing network and three fifths of a flat one on arrival and takes every flat or shrinking network within five years, and the only coexistence-shaped outcome is private supply in a growing network; what holds it is the investment decision, since its economics are project economics. A third design, a DRAM board with the hottest half of the dataset in on-board SRAM, sits between the two: it coexists only in a growing network at cheap GPU electricity and fails the cost conditions in a flat or shrinking one, and the dataset's size floor is a real lever on it where it was none on the SRAM die. What holds the die is the investment decision, not the hash; every chip figure here is modelled, not measured, and the chip's hardware cost per unit of work is approximate within 2x. | The litepaper (the chip model) | MODELLED; designed (the bracket modelled; the GPU side tested by the team) | `docs/design/class-v6-rotating-family.md` section 10 (10.0h to 10.0n, 8 October 2026); `docs/plans/igneum-2.0.md` D3 and D4 | the scoring rule in the close (the minimum over workloads of the maximum over free adversarial designs of the GPU's joules per hash over the adversary's, under the 10 percent GPU-cost budget at the lock, the verifier limit, cross-vendor correctness and hardware accessibility); the placed rows are D3's | the GPU side measured: the RTX 5080 at its 1,100 MHz lock 2.06 microjoules per hash and the RTX 5090 at its 1,300 MHz lock 2.33 (8 October 2026, the project's own rigs and rented pods); the chip side synthesised and claimed, its placed gated row pending | none yet | | 7 | Mining and proving together on one card needs a 16 GB card at the 5.5 GiB dataset floor: the miner holds about 6.1 GiB and a compressed shard proof peaks at about 7.5 GiB, so 8 GB and 12 GB cards time-share (the app pauses the miner for the proof). NVIDIA proves; AMD and Apple mine | The litepaper (Proving, vs RandomX); the miner page | TEAM-REPORTED; tested by the team | `docs/analysis/class-v6/coexist-rows.md` (the 5.5 GiB ds55 miner beside igneum-prove-host-0317, compressed at threshold 2^26, the served sm_86 and sm_89 floors) | the RESULT rows in that file, verbatim from the runs | rented RTX 3060 12 GB: 26.82 MH/s at 117.4 W with 6,129 MiB resident; the compressed shard 13.2 s, peak 7,525 MiB (6,129 + 7,525 = 13,654 MiB against 12,288); rented RTX 4060 8 GB: 18.84 MH/s, 6,116 MiB resident; the shard 8.2 s, peak 7,532 MiB; 8 October 2026 | none yet | ## Count by status diff --git a/site/evidence.html b/site/evidence.html index 922e018bf..2d2578aa8 100644 --- a/site/evidence.html +++ b/site/evidence.html @@ -272,7 +272,7 @@ td.mono{font-family:var(--f-mono);font-size:12.5px;min-width:180px}td.iv{color:v 3The 64-register window per lane costs a GPU under 1 percent of rate at stock, and at most 5 percent per load with the liveness chain
The litepaper (class v6); docs/plans/igneum-2.0.md D1 (the placed 64-register rows)
TEAM-REPORTED; tested by the teamdocs/analysis/class-v6/connected-state.md section 4; docs/design/class-v6-rotating-family.md section 10.0eThe class v5 nvcc harness and the kit worker, both packs on the same card minutes apart, 250 batches of 2^24, nvidia-smi at 1 Hz, vectors PASS on every row (connected-state.md section 4); the per-load rows of the full chain against the base (class-v6-rotating-family.md 10.0e)8 October 2026, rented RTX 5090 (575 W cap) and RTX 4090 (450 W cap) at stock: energy per hash +0.6 percent on the RTX 5090 and -0.9 percent on the RTX 4090, inside the run-to-run noise; under 1 percent of rate; 80 to 87 registers per thread, no spill (all measured). Per load: RTX 5090 16.7 nJ base, 17.6 nJ full chain; RTX 4090 26.0 nJ, 27.0 nJ (measured). The lock row on the project's own rigs is owednone yet 4Reorganising the same work around live state (the connected-state variant, experiment D2(a)) does not reduce a specialised chip's edge: KILL as a class
docs/plans/igneum-2.0.md D2(a); this page
TEAM-REPORTED; tested by the team (a published failure)docs/analysis/class-v6/connected-state.md (the verdict, section 6)The census, liveness and GPU rows in connected-state.md sections 2 to 4; the chip side priced on the drawn program by synthesis (a model, never a lower bound)8 October 2026, verdict 17:25 UK: the window is necessary (63 of 64 registers live at every address, measured) but only its width reaches the chip, +1.2 pJ per lane-op at N5 (synthesised); the window moves the chip's edge 1.10x node for node against a 1.25x gate (modelled); the GPU side +0.6 percent energy per hash on the RTX 5090, -0.9 percent on the RTX 4090 at stock (measured). Rearranging the dependency graph of the same operations moves neither sidenone yet 5"A GPU-secured network for Ethereum-compatible applications and verifiable computation." served on every page
Every page
PROPOSED; designed (served)docs/plans/igneum-2.0.md (the objective: the positioning line)node tools/ci/ledger-text-check.mjs: the sentence pinned (R0) on the home page, the litepaper and this page8 October 2026: on this page; the home page and the litepaper carry it as their 2.0 text lands (designed)none yet -6The chip claim as served: "Igneum remains competitive on accessible commodity GPUs even when specialised mining hardware is assumed to exist, remain compatible and seek profit; its security does not rely on identifying that hardware or retiring it through emergency changes." Under it the three statements, separate: energy ("Current modelling estimates a 1.5x to 3.1x energy-efficiency advantage for the specialised designs assessed as complete machines against the GPU tier, from a board on commodity DRAM at 1.5x to an SRAM-store die at 3.1x (1.5x to 2.3x on the GPU's own node)." Per machine on the same node and a node ahead: the DRAM board 1.5x and 1.8x, the hybrid 1.9x and 2.4x, the die 2.3x and 3.1x; 3x to 6x per dollar of hardware at list price. MODELLED: the GPU side measured, the chip's core placed and routed, the rest of the machine modelled, no chip measured, hardware cost approximate within 2x), economic and response capability, rotation an optional improvement. Class v5 derives the dataset from chain state; whether that excludes a specialised design is under evaluation (Deliverable 3), since a design that tracks state is not excluded by staleness. The energy ratio is not the pass criterion: the coexistence model ([docs/analysis/class-v6/coexistence-model.md](https://git.igneum.network/igneum-network/igneum/src/branch/master/docs/analysis/class-v6/coexistence-model.md)) is, and its first run's result is served with its conditions: A specialised supplier may earn a normal return; ordinary GPUs remain sufficiently close in total cost, widely obtainable and useful outside mining that new operators can still compete. On today's modelled rows that holds for the chip anyone can build, a stored-dataset board on commodity DRAM, at a productive life of one to three years: its cost per accepted unit of work sits within the range of the best GPU owner and entrant, it passes six of the seven conditions (a third of the network in boards now costs less than a year's revenue at the final hardware price), and a fleet of it holds a minority of the network with GPU entrants still setting the price. For the SRAM-store die the outcome turns on its hardware cost per unit of work, which the two models price five times apart: at the dearer figure it behaves like the hybrid board and coexists only in a growing network, at the cheaper one a modest fleet takes most of the network once the die exists with its development paid; at either figure it fails the fleet-cost and margin conditions in a flat or shrinking network, and its energy advantage never decides it. A third design, a DRAM board with the hottest half of the dataset in on-board SRAM, sits between the two: it coexists only in a growing network at cheap GPU electricity and fails the cost conditions in a flat or shrinking one, and the dataset's size floor is a real lever on it where it was none on the SRAM die. What holds the die is the investment decision, not the hash; every chip figure here is modelled, not measured, and the chip's hardware cost per unit of work is approximate within 2x.
The litepaper (the chip model)
MODELLED; designed (the bracket modelled; the GPU side tested by the team)docs/design/class-v6-rotating-family.md section 10 (10.0h to 10.0n, 8 October 2026); docs/plans/igneum-2.0.md D3 and D4the scoring rule in the close (the minimum over workloads of the maximum over free adversarial designs of the GPU's joules per hash over the adversary's, under the 10 percent GPU-cost budget at the lock, the verifier limit, cross-vendor correctness and hardware accessibility); the placed rows are D3'sthe GPU side measured: the RTX 5080 at its 1,100 MHz lock 2.06 microjoules per hash and the RTX 5090 at its 1,300 MHz lock 2.33 (8 October 2026, the project's own rigs and rented pods); the chip side synthesised and claimed, its placed gated row pendingnone yet +6The chip claim as served: "Igneum remains competitive on accessible commodity GPUs even when specialised mining hardware is assumed to exist, remain compatible and seek profit; its security does not rely on identifying that hardware or retiring it through emergency changes." Under it the three statements, separate: energy ("Current modelling estimates a 1.5x to 3.1x energy-efficiency advantage for the specialised designs assessed as complete machines against the GPU tier, from a board on commodity DRAM at 1.5x to an SRAM-store die at 3.1x (1.5x to 2.3x on the GPU's own node)." Per machine on the same node and a node ahead: the DRAM board 1.5x and 1.8x, the hybrid 1.9x and 2.4x, the die 2.3x and 3.1x; 3x to 6x per dollar of hardware at list price. MODELLED: the GPU side measured, the chip's core placed and routed, the rest of the machine modelled, no chip measured, hardware cost approximate within 2x), economic and response capability, rotation an optional improvement. Class v5 derives the dataset from chain state; whether that excludes a specialised design is under evaluation (Deliverable 3), since a design that tracks state is not excluded by staleness. The energy ratio is not the pass criterion: the coexistence model ([docs/analysis/class-v6/coexistence-model.md](https://git.igneum.network/igneum-network/igneum/src/branch/master/docs/analysis/class-v6/coexistence-model.md)) is, and its first run's result is served with its conditions: A specialised supplier may earn a normal return; ordinary GPUs remain sufficiently close in total cost, widely obtainable and useful outside mining that new operators can still compete. On today's modelled rows that holds for the chip anyone can build, a stored-dataset board on commodity DRAM, at a productive life of one to three years: its cost per accepted unit of work sits within the range of the best GPU owner and entrant, it passes six of the seven conditions (a third of the network in boards now costs less than a year's revenue at the final hardware price), and a fleet of it holds a minority of the network with GPU entrants still setting the price. For the SRAM-store die the outcome turns on its hardware cost per unit of work, not its energy advantage: at the reconciled machine cost, set by the power train and the shadow core rather than the die, it fails the cost, hardware, fleet and margin conditions at every point of the band, a modest fleet holds about two fifths of a growing network and three fifths of a flat one on arrival and takes every flat or shrinking network within five years, and the only coexistence-shaped outcome is private supply in a growing network; what holds it is the investment decision, since its economics are project economics. A third design, a DRAM board with the hottest half of the dataset in on-board SRAM, sits between the two: it coexists only in a growing network at cheap GPU electricity and fails the cost conditions in a flat or shrinking one, and the dataset's size floor is a real lever on it where it was none on the SRAM die. What holds the die is the investment decision, not the hash; every chip figure here is modelled, not measured, and the chip's hardware cost per unit of work is approximate within 2x.
The litepaper (the chip model)
MODELLED; designed (the bracket modelled; the GPU side tested by the team)docs/design/class-v6-rotating-family.md section 10 (10.0h to 10.0n, 8 October 2026); docs/plans/igneum-2.0.md D3 and D4the scoring rule in the close (the minimum over workloads of the maximum over free adversarial designs of the GPU's joules per hash over the adversary's, under the 10 percent GPU-cost budget at the lock, the verifier limit, cross-vendor correctness and hardware accessibility); the placed rows are D3'sthe GPU side measured: the RTX 5080 at its 1,100 MHz lock 2.06 microjoules per hash and the RTX 5090 at its 1,300 MHz lock 2.33 (8 October 2026, the project's own rigs and rented pods); the chip side synthesised and claimed, its placed gated row pendingnone yet 7Mining and proving together on one card needs a 16 GB card at the 5.5 GiB dataset floor: the miner holds about 6.1 GiB and a compressed shard proof peaks at about 7.5 GiB, so 8 GB and 12 GB cards time-share (the app pauses the miner for the proof). NVIDIA proves; AMD and Apple mine
The litepaper (Proving, vs RandomX); the miner page
TEAM-REPORTED; tested by the teamdocs/analysis/class-v6/coexist-rows.md (the 5.5 GiB ds55 miner beside igneum-prove-host-0317, compressed at threshold 2^26, the served sm_86 and sm_89 floors)the RESULT rows in that file, verbatim from the runsrented RTX 3060 12 GB: 26.82 MH/s at 117.4 W with 6,129 MiB resident; the compressed shard 13.2 s, peak 7,525 MiB (6,129 + 7,525 = 13,654 MiB against 12,288); rented RTX 4060 8 GB: 18.84 MH/s, 6,116 MiB resident; the shard 8.2 s, peak 7,532 MiB; 8 October 2026none yet

Click a column heading to sort; click again to reverse. Versions: the Igneum 2.0 devnet (igneum-devnet-4) runs since its first block at 17:14 UK on 8 October 2026; the release manifest, machine-readable, is regenerated at its first block and fills these at build time. As last read it names node 4cdcc488 on release-2.0.0-node, igneum-pow 1c420786, chain id 4465, read 8 October 2026, 17:2x UK. The labels stay distinct: a claim never moves up a label without the artefact the next table names. The public reference repository exists (the specifications, the pow crate, the simulators and the test material, at git.igneum.network); the full node, the miner and the proving code open later, so a row that cites them is tested by the team at most until they do. Node fork commits are the node fork’s; a document path is the repository’s. The source of this page is docs/evidence.md in the repository.

diff --git a/site/index.html b/site/index.html index 25788a660..222fd3b9e 100644 --- a/site/index.html +++ b/site/index.html @@ -263,7 +263,7 @@
01
The same card finds the block and proves it.

Both jobs pay. When you stop, the card still games.

02
A fair start.

Nobody holds a coin before block one. The protocol carries no fee. The one payment to the project is the Ember software’s optional 1% dev fee, like other GPU miners, off with one flag.

-
03
Built for graphics cards.

Igneum remains competitive on accessible commodity GPUs even when specialised mining hardware is assumed to exist, remain compatible and seek profit; its security does not rely on identifying that hardware or retiring it through emergency changes. Class v6 adopts the 64-register window and retains it across every rotation. Current modelling estimates a 1.5x to 3.1x energy-efficiency advantage for the specialised designs assessed as complete machines against the GPU tier, from a board on commodity DRAM at 1.5x to an SRAM-store die at 3.1x (1.5x to 2.3x on the GPU's own node). The long-program and select-tree proposals were rejected. Economic resistance depends on development cost, deployment economics and productive hardware lifetime; family transitions receive an obsolescence benefit only where a loss of competitiveness is demonstrated; programmable multi-epoch designs are included in the assessment. MODELLED: the GPU side measured (the RTX 5090 at its lock, 8 October 2026), the chip’s core placed and routed, the rest of the machine modelled, no chip measured, hardware cost approximate within 2x; the coexistence model finds the DRAM board passes six of seven conditions, the hybrid coexists only in a growing chain at cheap GPU electricity, and the SRAM die’s verdict turns on its hardware ticket; a 5090 locked at its knee pays 82 W for the class v4 shadow work (measured, 7 October 2026). Every number with its label, the harness and the scoring rules.

+
03
Built for graphics cards.

Igneum remains competitive on accessible commodity GPUs even when specialised mining hardware is assumed to exist, remain compatible and seek profit; its security does not rely on identifying that hardware or retiring it through emergency changes. Class v6 adopts the 64-register window and retains it across every rotation. Current modelling estimates a 1.5x to 3.1x energy-efficiency advantage for the specialised designs assessed as complete machines against the GPU tier, from a board on commodity DRAM at 1.5x to an SRAM-store die at 3.1x (1.5x to 2.3x on the GPU's own node). The long-program and select-tree proposals were rejected. Economic resistance depends on development cost, deployment economics and productive hardware lifetime; family transitions receive an obsolescence benefit only where a loss of competitiveness is demonstrated; programmable multi-epoch designs are included in the assessment. MODELLED: the GPU side measured (the RTX 5090 at its lock, 8 October 2026), the chip’s core placed and routed, the rest of the machine modelled, no chip measured, hardware cost approximate within 2x; the coexistence model finds the DRAM board passes six of seven conditions, the hybrid coexists only in a growing chain at cheap GPU electricity, and the SRAM die fails four conditions at its reconciled machine cost; a 5090 locked at its knee pays 82 W for the class v4 shadow work (measured, 7 October 2026). Every number with its label, the harness and the scoring rules.

diff --git a/site/litepaper.html b/site/litepaper.html index a80b9bfe0..1fcade13b 100644 --- a/site/litepaper.html +++ b/site/litepaper.html @@ -467,11 +467,11 @@ body.all .pager{display:none}

The work that waits can grow. Class v4 adds a block of latency-shadow arithmetic to every hash, about 100,000 integer operations that run while the memory reads are in flight, so a chip that stores the whole dataset still has to pay for a core. That size sits on a ladder fixed at genesis, six rungs from about 100,000 to about 1,000,000 operations, and it moves one rung at a time only when 90 percent of blue blocks in each of seven consecutive days ask for it; it can never move two rungs inside a week and never past a rung the reference verifier cannot check under 10 ms with its sibling thread busy (measured on the build server, 6 October 2026: the first three rungs pass at 8.8, 8.9 and 9.2 ms, the fourth misses by 0.08 ms on a loaded box and stays out until a quiet re-measurement, the two doublings are out at 12.4 and 15.0 ms). What it buys against a chip is scored under the rule in the chip model below. What it costs, per rung, is measured too: the Apple tier gives up 3 points of rate at the first step and 6 more at the second, the RTX 5090 nothing until the second; so the miners who pay for a step are the ones who take it.

The chip model

Igneum remains competitive on accessible commodity GPUs even when specialised mining hardware is assumed to exist, remain compatible and seek profit; its security does not rely on identifying that hardware or retiring it through emergency changes. Class v6 adopts the 64-register window and retains it across every rotation. Current modelling estimates a 1.5x to 3.1x energy-efficiency advantage for the specialised designs assessed as complete machines against the GPU tier, from a board on commodity DRAM at 1.5x to an SRAM-store die at 3.1x (1.5x to 2.3x on the GPU's own node). The long-program and select-tree proposals were rejected. Economic resistance depends on development cost, deployment economics and productive hardware lifetime; family transitions receive an obsolescence benefit only where a loss of competitiveness is demonstrated; programmable multi-epoch designs are included in the assessment.

-

The labels. MODELLED. The GPU side is measured (the RTX 5090 at its 1,300 MHz lock on class v4, 8 October 2026). The chip’s core is placed and routed on ASAP7 with SRAM macros and scaled on claimed node factors; the chip’s memory and the rest of the machine (controller, host share, PSU, VRM, cooling) are modelled; no chip is measured, and the chip’s hardware cost is approximate within 2x. Beside it the coexistence verdict at these energies: the DRAM board passes six of the model’s seven conditions, the hybrid coexists only in a growing chain at cheap GPU electricity, the SRAM die’s verdict turns on its hardware ticket, which two sources price five times apart (reconciliation owed). The k lane’s own placed row amends the figures if it differs. The GPU side is measured: an RTX 5080 at its 1,100 MHz core lock, 2.06 microjoules per hash, and an RTX 5090 at its 1,300 MHz lock, 2.33 microjoules per hash, both under class v4, on the project’s own rigs and rented pods, 8 October 2026; the card’s cost of the window is measured too (a rented RTX 5090 and RTX 4090 at stock, 8 October 2026: within 5 percent per load with the liveness chain, no register spill). The chip side is synthesised and claimed: the clock-gated sequencer core with the 64-register window on ASAP7, scaled to N3 on the foundry’s headline factors (k about 0.37 at N3 and 0.51 node for node for the base core, the gated window adding about 0.13 of k against an adversary with a flop register file; the window’s liveness measured at 61 of 64 values necessary, its cost to the card measured under 5 percent); the window’s k is synthesis-derived and not a lower bound, and the multi-family adversary lane’s first core (its state in a macro) reads the window’s defence as close to nothing, a disagreement between two models that the placed rows settle. The chip’s memory is modelled: the GDDR7 board of the chip model. The placed gated figure is expected near 2.6x to 3.1x a node ahead and 2.3x to 2.7x node for node (approximate) and is served when its row lands.

+

The labels. MODELLED. The GPU side is measured (the RTX 5090 at its 1,300 MHz lock on class v4, 8 October 2026). The chip’s core is placed and routed on ASAP7 with SRAM macros and scaled on claimed node factors; the chip’s memory and the rest of the machine (controller, host share, PSU, VRM, cooling) are modelled; no chip is measured, and the chip’s hardware cost is approximate within 2x. Beside it the coexistence verdict at these energies: the DRAM board passes six of the model’s seven conditions, the hybrid coexists only in a growing chain at cheap GPU electricity, the SRAM die fails the cost, hardware, fleet and margin conditions at its reconciled machine cost, and only private supply in a growing network coexists. The k lane’s own placed row amends the figures if it differs. The GPU side is measured: an RTX 5080 at its 1,100 MHz core lock, 2.06 microjoules per hash, and an RTX 5090 at its 1,300 MHz lock, 2.33 microjoules per hash, both under class v4, on the project’s own rigs and rented pods, 8 October 2026; the card’s cost of the window is measured too (a rented RTX 5090 and RTX 4090 at stock, 8 October 2026: within 5 percent per load with the liveness chain, no register spill). The chip side is synthesised and claimed: the clock-gated sequencer core with the 64-register window on ASAP7, scaled to N3 on the foundry’s headline factors (k about 0.37 at N3 and 0.51 node for node for the base core, the gated window adding about 0.13 of k against an adversary with a flop register file; the window’s liveness measured at 61 of 64 values necessary, its cost to the card measured under 5 percent); the window’s k is synthesis-derived and not a lower bound, and the multi-family adversary lane’s first core (its state in a macro) reads the window’s defence as close to nothing, a disagreement between two models that the placed rows settle. The chip’s memory is modelled: the GDDR7 board of the chip model. The placed gated figure is expected near 2.6x to 3.1x a node ahead and 2.3x to 2.7x node for node (approximate) and is served when its row lands.

Three statements, kept separate. The baseline is the hash as it stands under the scoring rule; rotation is an optional improvement to that baseline, not the mechanism the claim rests on.

- +
StatementWhat it saysLabel and date
Energy resistancePer machine against a locked RTX 5090, node for node and a node ahead: the DRAM board 1.5x (1.3x to 1.7x) and 1.8x; the board with SRAM holding the hottest half of the dataset 1.9x and 2.4x; the SRAM-store die 2.3x and 3.1x; per dollar of hardware at list price 3x to 6x (modelled on the placed full 18-family core, routed 8 October 2026). The honest tier moves to the next node with every GPU generation; a chip must tape out again. Whole machine per tier (synthesis with the SRAM band and node factors, claimed; placed rows to follow): the complete GDDR7 machine about 1.8x the RTX 5090 at its lock per joule node for node and 2.1x a node ahead; 1.6x and 1.9x the RTX 5080; 2.8x and 3.3x the Ada, Ampere and RX 9070 XT cohort; about 1.5x the Apple tier, reported, never headlined.MODELLED: placed and routed core energies against the measured RTX 5090 lock row, 8 October 2026, no chip measured
Economic resistanceWhether a chip gets built depends on development cost, deployment economics and productive hardware lifetime. The first cut of the profitability surface: the price at which a project pays scales as the project cost over its share of the chain times its discounted life, and moves by under 5 percent with the per-joule edge; a fixed-lane chip under rotation needs 4x the price a programmable one needs. No threshold is the headline: the five-year coexistence model (Deliverable 4; its first run is docs/analysis/class-v6/coexistence-model.md, every row modelled) replaces any capex wall: the DRAM board passes six of its seven success conditions at a one to three year life; the SRAM die's verdict turns on its hardware ticket; the larger half of a chip's edge is capital cost per accepted hash, not joules. The result, as the model words it: A specialised supplier may earn a normal return; ordinary GPUs remain sufficiently close in total cost, widely obtainable and useful outside mining that new operators can still compete. On today's modelled rows that holds for the chip anyone can build, a stored-dataset board on commodity DRAM, at a productive life of one to three years: its cost per accepted unit of work sits within the range of the best GPU owner and entrant, it passes six of the seven conditions (a third of the network in boards now costs less than a year's revenue at the final hardware price), and a fleet of it holds a minority of the network with GPU entrants still setting the price. For the SRAM-store die the outcome turns on its hardware cost per unit of work, which the two models price five times apart: at the dearer figure it behaves like the hybrid board and coexists only in a growing network, at the cheaper one a modest fleet takes most of the network once the die exists with its development paid; at either figure it fails the fleet-cost and margin conditions in a flat or shrinking network, and its energy advantage never decides it. A third design, a DRAM board with the hottest half of the dataset in on-board SRAM, sits between the two: it coexists only in a growing network at cheap GPU electricity and fails the cost conditions in a flat or shrinking one, and the dataset's size floor is a real lever on it where it was none on the SRAM die. What holds the die is the investment decision, not the hash; every chip figure here is modelled, not measured, and the chip's hardware cost per unit of work is approximate within 2x. The model holds under every market structure for the DRAM board (private supply, hardware sales, multiple suppliers, with no single supplier above a quarter of the network), and the SRAM die under none but private supply in a growing network; the thresholds live in the model's sensitivity workbook (docs/analysis/class-v6/coexistence-workbook.md), never on a page.MODELLED, the coexistence model's first run, 8 October 2026
Economic resistanceWhether a chip gets built depends on development cost, deployment economics and productive hardware lifetime. The first cut of the profitability surface: the price at which a project pays scales as the project cost over its share of the chain times its discounted life, and moves by under 5 percent with the per-joule edge; a fixed-lane chip under rotation needs 4x the price a programmable one needs. No threshold is the headline: the five-year coexistence model (Deliverable 4; its first run is docs/analysis/class-v6/coexistence-model.md, every row modelled) replaces any capex wall: the DRAM board passes six of its seven success conditions at a one to three year life; the SRAM die fails four conditions at its reconciled machine cost; the larger half of a chip's edge is capital cost per accepted hash, not joules. The result, as the model words it: A specialised supplier may earn a normal return; ordinary GPUs remain sufficiently close in total cost, widely obtainable and useful outside mining that new operators can still compete. On today's modelled rows that holds for the chip anyone can build, a stored-dataset board on commodity DRAM, at a productive life of one to three years: its cost per accepted unit of work sits within the range of the best GPU owner and entrant, it passes six of the seven conditions (a third of the network in boards now costs less than a year's revenue at the final hardware price), and a fleet of it holds a minority of the network with GPU entrants still setting the price. For the SRAM-store die the outcome turns on its hardware cost per unit of work, not its energy advantage: at the reconciled machine cost, set by the power train and the shadow core rather than the die, it fails the cost, hardware, fleet and margin conditions at every point of the band, a modest fleet holds about two fifths of a growing network and three fifths of a flat one on arrival and takes every flat or shrinking network within five years, and the only coexistence-shaped outcome is private supply in a growing network; what holds it is the investment decision, since its economics are project economics. A third design, a DRAM board with the hottest half of the dataset in on-board SRAM, sits between the two: it coexists only in a growing network at cheap GPU electricity and fails the cost conditions in a flat or shrinking one, and the dataset's size floor is a real lever on it where it was none on the SRAM die. What holds the die is the investment decision, not the hash; every chip figure here is modelled, not measured, and the chip's hardware cost per unit of work is approximate within 2x. The model holds under every market structure for the DRAM board (private supply, hardware sales, multiple suppliers, with no single supplier above a quarter of the network), and the SRAM die under none but private supply in a growing network; the thresholds live in the model's sensitivity workbook (docs/analysis/class-v6/coexistence-workbook.md), never on a page.MODELLED, the coexistence model's first run, 8 October 2026
Response capabilityRotation is an optional improvement, not the mechanism. A passed rotation boundary proves the rotation works, not that hardware dies. The schedule: a new program every hour, a parameter era every week, a family epoch every 180 days, an emergency vote when miners call one. Rotation costs a chip versatility, not life: the family bank is firmware plus about 43 percent of core cells, and no transition carries an obsolescence credit (modelled).measured per boundary, 8 October 2026; the family-bank cost modelled, 8 October 2026

What a miner sees from this. Class v4 costs a 5090 145 W more unlocked, 88 W more at a 1,400 MHz core lock and 82 W at the best operating points (class v4 at 1,200 MHz, class v3 at 1,300; the knee is 1,300 MHz on both), for 0.2 percent more rate (measured, 7 October 2026; the 80 W read on 6 October was at the app's tuned cap); an M5 Max 16 W more for 1.5 percent, an RX 9070 XT and an RTX 4070 nothing (measured, 6 October 2026). The ladder that sets how much work rides in the shadow starts at rung 0 at genesis and climbs by miner signal; its third rung is inadmissible today because a server core verifies it in 10.85 ms, over the gate (measured, 7 October 2026). The devnet runs class v4 from its first block. Classes rotate on findings and at least yearly; a class change is a release activated by block height. Class v5 is built to cross by height; its dataset keyed by the chain's own state is under evaluation (Deliverable 3) and is not counted as a defence until justified. Class v6 is the design in progress (opened 8 October 2026), with four layers as its spine: per-era draws of the parameters a release now fixes, a dataset whose size tracks the chain state (under evaluation, Deliverable 3), scheduled family epochs by height, and the acceptance floor generalised to every era’s draw. The next test of the model is an internal adversarial pass, not an independent review: three lanes that have never worked on the hash code attack the mixer, the chained cache and the acceptance rule with only what an outsider has (the public kit, the frozen object, the spec, the harnesses) and publish the break or the bound they reach. No outside review has run yet.