From e8eb95760918fc28b169c9c8618ec20e19659697 Mon Sep 17 00:00:00 2001
From: igneum-labs <337424239+igneum-labs@users.noreply.github.com>
Date: Tue, 6 Oct 2026 19:17:57 +0000
Subject: [PATCH] Site: the miner page's proving claim matches the evidence
 (row 16), the h1 is 42 characters, lever 4 says shipped in 0.3.6

Audit fixes, copy only. (1) "every NVIDIA card from 8 GB proves" cited docs/analysis/prover-tiers-real-cards.md, which
is not in the repository, and contradicted evidence row 16; the hero lead, the proving feat, the three meta descriptions
and the homepage lead and h2 now say what row 16 and docs/analysis/amd-proving.md state: measured on an RTX 5090 with the
shipped SP1 server (13.9 GB floor), a 32 GB card mines and proves, a 24 GB card proves the full shard alone (from the
5090's allocation, not yet run on a 24 GB card), 16 GB empty shards only, 12 GB and under nothing on this build; the
eleven rented cards of 6 October are "measured, record pending". Each links /evidence#claims. (2) The h1 was 96
characters and seven lines on a phone; it is "Install. Start. The card mines and proves." (42) with the tiers in the lead.
(3) Lever row 4 said "Ships in 0.3.6" at 0.3.14; it now says shipped in 0.3.6 (the miner-latency merge, release-0.3.6.md
section 2), the app at 0.3.14.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
---
 site/index.html |  6 +++---
 site/miner.html | 14 +++++++-------
 2 files changed, 10 insertions(+), 10 deletions(-)

diff --git a/site/index.html b/site/index.html
index c8e0ec59..46ed9909 100644
--- a/site/index.html
+++ b/site/index.html
@@ -342,7 +342,7 @@ pre{margin:0;font-family:var(--f-mono);font-size:13px;line-height:1.6;color:var(
     <div class="hero-copy">
       <div class="eyebrow ember">GPUs are back · for good</div>
       <h1>Mined by GPUs.<br>Proven by fire.</h1>
-      <p class="lead">Built for graphics cards. A new mining program every hour, compiled on the card. In our public model the strongest chip reaches 5x to 9x per joule against an RTX 5090 today, about 2x once the lever now in its gates ships: <a href="/bench#counter-asic-2-0-the-numbers" style="color:var(--molten)">the numbers</a>. Every NVIDIA card from 8 GB proves and is paid for it; 12 GB and up mine and prove; AMD and Apple cards mine. No premine, no stake, no foundation.</p>
+      <p class="lead">Built for graphics cards. A new mining program every hour, compiled on the card. In our public model the strongest chip reaches 5x to 9x per joule against an RTX 5090 today, about 2x once the lever now in its gates ships: <a href="/bench#counter-asic-2-0-the-numbers" style="color:var(--molten)">the numbers</a>. Every card mines; a 32 GB NVIDIA card mines and proves, a 24 GB card proves a full shard alone, measured on an RTX 5090 (<a href="/evidence#claims" style="color:var(--molten)">evidence, row 16</a>); AMD and Apple cards mine. No premine, no stake, no foundation.</p>
       <div class="cta">
         <a href="#mine" class="btn primary">See the miner</a>
         <a href="/litepaper" class="btn">Read the litepaper</a>
@@ -491,7 +491,7 @@ pre{margin:0;font-family:var(--f-mono);font-size:13px;line-height:1.6;color:var(
 <section id="mine" class="light">
   <div class="wrap">
     <div class="sec-head reveal">
-      <h2>One click. The card mines; every NVIDIA card from 8 GB proves, 12 GB and up mine and prove.</h2>
+      <h2>One click. The card mines; a 24 GB NVIDIA card proves as well.</h2>
       <p><span data-product="full">Igneum Ember</span> finds your GPU, makes a wallet for you and runs the node, the miner and the prover as one app.</p>
     </div>
     <div class="split">
@@ -676,7 +676,7 @@ pre{margin:0;font-family:var(--f-mono);font-size:13px;line-height:1.6;color:var(
 <!-- footer:end -->
 
 <!-- journey:start -->
-<script type="application/json" id="journey-data">{"updated":"2026-10-06","stage":"phase-3","phases":[{"id":"phase-1","name":"Specification","when":"Oct to Nov 2026","status":"active","line":"Mining generator, shard proving, finality rules, written for external review"},{"id":"phase-2","name":"Prove the proving","when":"Nov 2026 to Jan 2027","status":"active","line":"Mining program proven on Apple, NVIDIA and AMD; shard proving benchmark on consumer cards still to run","gate":"A fixed published workload, job received to accepted proof on a declared consumer card, sustained with no growing backlog, reproduced by three independent operators"},{"id":"phase-3","name":"Devnet","when":"Started 3 Oct 2026, 20 nodes by Mar 2027","status":"active","line":"BlockDAG node mining on the new program, GPU miners on three vendors, EVM execution in build","gate":"1 block a second held with proofs under 60 s behind the tip"},{"id":"phase-4","name":"Finality and job market","when":"Apr to Jul 2027","status":"next","line":"Sustained-mining finality, external proving jobs, miner client with auto-switching","gate":"Finality design passes external review and one rollup signs for testnet"},{"id":"phase-5","name":"Public testnet","when":"Aug to Oct 2027","status":"next","line":"One-click miner app on Windows, macOS and Linux, HiveOS, pools, the first rollup as a proving customer, no coin yet","gate":"1,000 independent miners run 30 days and rollup proofs are delivered on time"},{"id":"phase-6","name":"Mainnet fair launch","when":"Nov 2027","status":"next","line":"Genesis with no premine, 30-day ramp. No listing is arranged, promised or sought by the project"}],"log":[{"date":"2026-10-06","text":"Counter ASIC 3.0 item 2: the per-day derivation","short":"Counter ASIC 3.0 item 2"},{"date":"2026-10-06","text":"Counter ASIC 3.0 item 8: program work in the latency shadow","short":"Counter ASIC 3.0 item 8"},{"date":"2026-10-06","text":"Counter ASIC 3.0 item 6: the reserve families' step costs","short":"Counter ASIC 3.0 item 6"},{"date":"2026-10-06","text":"Counter ASIC 3.0 gate run, the hash side","short":"Counter ASIC 3.0 gate run, the hash side"},{"date":"2026-10-06","text":"16:01Z: Ember run 6 on PC 1","short":"16:01Z: Ember run 6 on PC 1"},{"date":"2026-10-06","text":"00:4xZ, the empty `/api/state` reply","short":"00:4xZ, the empty `/api/state` reply"},{"date":"2026-10-06","text":"07:12Z to 07:17Z, the host's chain mode with --save-shards records and --prev, on the Apple M5 Max's CPU","short":"07:12Z to 07:17Z, the host's chain mode with --save-shards records…"},{"date":"2026-10-06","text":"07:52Z to 08:24Z, the segment-aligned prover beside the miner on PC 2's RTX 5090","short":"07:52Z to 08:24Z, the segment-aligned prover beside the miner on PC…"},{"date":"2026-10-06","text":"08:26Z to 08:35Z, the fast-time harness on the fresh-record rule","short":"08:26Z to 08:35Z, the fast-time harness on the fresh-record rule"},{"date":"2026-10-06","text":"12:25 to 13:20Z, the finality route: why 26 fresh nodes lost the seed every checkpoint","short":"12:25 to 13:20Z, the finality route: why 26 fresh nodes lost the…"},{"date":"2026-10-05","text":"The 9070 XT on the eGPU: why 17.9 MH/s, and what moved","short":"The 9070 XT on the eGPU"},{"date":"2026-10-05","text":"Ember Tune: the two-knob efficiency tune, the fleet prior, and what PC 1 could measure tonight","short":"Ember Tune: the two-knob efficiency tune, the fleet prior"},{"date":"2026-10-05","text":"Read width of the lottery hash: 4, 16 and 64-byte loads, a per-load mix, a written scratch; three cards","short":"Read width of the lottery hash"},{"date":"2026-10-05","text":"The hot table on the M5 Max: a second table sized to GPU cache beside the 1 GiB dataset","short":"The hot table on the M5 Max"},{"date":"2026-10-05","text":"Mixer x4 and the cache growth rule: the class v3 dataset construction, with the x8 candidate","short":"Mixer x4 and the cache growth rule"},{"date":"2026-10-05","text":"The SP1 CPU prover on PC 1 beside the miners, and the backend survey: no zkVM proves on AMD","short":"The SP1 CPU prover on PC 1 beside the miners, and the backend survey"},{"date":"2026-10-05","text":"Proving v1: segment records, the chain rule, the unproven rule; what was measured tonight","short":"Proving v1: segment records, the chain rule"},{"date":"2026-10-05","text":"Dp4a-class throughput on the M5 Max: the dot4 emulation against the ALU chain","short":"Dp4a-class throughput on the M5 Max"},{"date":"2026-10-05","text":"Layer 3 scratch soundness","short":"Layer 3 scratch soundness"},{"date":"2026-10-05","text":"Epoch length as an era parameter : the Apple M5 Max's compile-ahead per program","short":"Epoch length as an era parameter"},{"date":"2026-10-05","text":"Aggregation cost on the RTX 5090: what a per-block aggregation spends and what each lever gives","short":"Aggregation cost on the RTX 5090"},{"date":"2026-10-05","text":"The C4 fix: certificate-driven reorg","short":"The C4 fix: certificate-driven reorg"},{"date":"2026-10-05","text":"EVM transaction relay: three nodes in a chain, every transaction sent to one end included by the other two miners","short":"EVM transaction relay"},{"date":"2026-10-05","text":"The prover carries both fee tables and the height switch: one pinned guest on either side of DAA 210,000","short":"The prover carries both fee tables and the height switch"},{"date":"2026-10-05","text":"FUD ledger sweep round 6","short":"FUD ledger sweep round 6"},{"date":"2026-10-05","text":"Live devnet: real transactions, the first non-empty shard proven and paid, and the exporter's block structure fixed","short":"Live devnet: real transactions, the first non-empty shard proven and…"},{"date":"2026-10-05","text":"The program id split: why the Apple M5 Max rejected PC 2's proofs, and the verifier at 114 s","short":"The program id split"},{"date":"2026-10-05","text":"Live devnet: the first shards proven, verified and paid","short":"Live devnet: the first shards proven, verified and paid"},{"date":"2026-10-04","text":"Sim/economy: mining versus proving under stress, agent-based","short":"Economy simulation: mining versus proving under stress"},{"date":"2026-10-04","text":"Difficulty rule under attack: pool hopping, pulsed rental, timestamp stretching, short-lane oscillation, epoch games, polluted window, block flood","short":"Difficulty rule attacked seven ways"},{"date":"2026-10-04","text":"Difficulty rule: timestamp attack fixed , simulator regression, 3-node forger test","short":"Timestamp attack on the difficulty rule fixed"},{"date":"2026-10-04","text":"Devnet-v4 integration: nine branches merged, 3-node test network on the merged node, Windows cross-build","short":"Devnet v4: nine branches merged into one node"},{"date":"2026-10-04","text":"Generator version 2 adopted: exact load count, fresh-source loads, program acceptance; every vector re-cut, three workers re-checked, 20,000-program census, devnet-v4 binaries rebuilt","short":"Generator v2 adopted: every hash does 128 distinct reads"},{"date":"2026-10-04","text":"Proving v0 on the RTX 5090: first GPU proof of an Igneum block","short":"First GPU proof of an Igneum block: 1.4 s on an RTX 5090"},{"date":"2026-10-04","text":"Devnet v4 cut-over: generator v2, 2/3 floor, three nodes and a seed on a fresh chain","short":"Devnet v4 live: generator v2, two-thirds floor, fresh chain"},{"date":"2026-10-04","text":"First hourly program swap on the live devnet: compile-ahead, no pause, two cards","short":"First live hourly swap: no pause on Mac, NVIDIA or AMD"},{"date":"2026-10-04","text":"Proving: devnet v4 shards on the Apple M5 Max CPU, loaded machine","short":"Proving: devnet v4 shards on the Apple M5 Max CPU, loaded machine"},{"date":"2026-10-04","text":"First finality lock on the live devnet: checkpoint 242 at 77.4% of all weight, two hours after genesis","short":"First live finality lock: 77.4% of weight, 17 voters"},{"date":"2026-10-04","text":"A node 60 s behind the clock is silently dead","short":"A node 60 s behind the clock is silently dead"},{"date":"2026-10-04","text":"First machine on the Igneum Miner app: PC 2's RTX 5090 at 118 MH/s, via Setup.exe","short":"First machine on the one-click app: a 5090 at 118 MH/s"}]}</script>
+<script type="application/json" id="journey-data">{"updated":"2026-10-06","stage":"phase-3","phases":[{"id":"phase-1","name":"Specification","when":"Oct to Nov 2026","status":"active","line":"Mining generator, shard proving, finality rules, written for external review"},{"id":"phase-2","name":"Prove the proving","when":"Nov 2026 to Jan 2027","status":"active","line":"Mining program proven on Apple, NVIDIA and AMD; shard proving benchmark on consumer cards still to run","gate":"A fixed published workload, job received to accepted proof on a declared consumer card, sustained with no growing backlog, reproduced by three independent operators"},{"id":"phase-3","name":"Devnet","when":"Started 3 Oct 2026, 20 nodes by Mar 2027","status":"active","line":"BlockDAG node mining on the new program, GPU miners on three vendors, EVM execution in build","gate":"1 block a second held with proofs under 60 s behind the tip"},{"id":"phase-4","name":"Finality and job market","when":"Apr to Jul 2027","status":"next","line":"Sustained-mining finality, external proving jobs, miner client with auto-switching","gate":"Finality design passes external review and one rollup signs for testnet"},{"id":"phase-5","name":"Public testnet","when":"Aug to Oct 2027","status":"next","line":"One-click miner app on Windows, macOS and Linux, HiveOS, pools, the first rollup as a proving customer, no coin yet","gate":"1,000 independent miners run 30 days and rollup proofs are delivered on time"},{"id":"phase-6","name":"Mainnet fair launch","when":"Nov 2027","status":"next","line":"Genesis with no premine, 30-day ramp. No listing is arranged, promised or sought by the project"}],"log":[{"date":"2026-10-06","text":"Counter ASIC 3.0 item 2: the per-day derivation","short":"Counter ASIC 3.0 item 2"},{"date":"2026-10-06","text":"Counter ASIC 3.0 item 8: program work in the latency shadow","short":"Counter ASIC 3.0 item 8"},{"date":"2026-10-06","text":"Counter ASIC 3.0 item 6: the reserve families' step costs","short":"Counter ASIC 3.0 item 6"},{"date":"2026-10-06","text":"Counter ASIC 3.0 gate run, the hash side","short":"Counter ASIC 3.0 gate run, the hash side"},{"date":"2026-10-06","text":"16:01Z: Ember run 6 on PC 1","short":"16:01Z: Ember run 6 on PC 1"},{"date":"2026-10-06","text":"00:4xZ, the empty `/api/state` reply","short":"00:4xZ, the empty `/api/state` reply"},{"date":"2026-10-06","text":"07:12Z to 07:17Z, the host's chain mode with --save-shards records and --prev, on the Apple M5 Max's CPU","short":"07:12Z to 07:17Z, the host's chain mode with --save-shards records…"},{"date":"2026-10-06","text":"07:52Z to 08:24Z, the segment-aligned prover beside the miner on PC 2's RTX 5090","short":"07:52Z to 08:24Z, the segment-aligned prover beside the miner on PC…"},{"date":"2026-10-06","text":"08:26Z to 08:35Z, the fast-time harness on the fresh-record rule","short":"08:26Z to 08:35Z, the fast-time harness on the fresh-record rule"},{"date":"2026-10-06","text":"12:25 to 13:20Z, the finality route: why 26 fresh nodes lost the seed every checkpoint","short":"12:25 to 13:20Z, the finality route: why 26 fresh nodes lost the…"},{"date":"2026-10-05","text":"The 9070 XT on the eGPU: why 17.9 MH/s, and what moved","short":"The 9070 XT on the eGPU"},{"date":"2026-10-05","text":"Ember Tune: the two-knob efficiency tune, the fleet prior, and what PC 1 could measure tonight","short":"Ember Tune: the two-knob efficiency tune, the fleet prior"},{"date":"2026-10-05","text":"Read width of the lottery hash: 4, 16 and 64-byte loads, a per-load mix, a written scratch; three cards","short":"Read width of the lottery hash"},{"date":"2026-10-05","text":"The hot table on the M5 Max: a second table sized to GPU cache beside the 1 GiB dataset","short":"The hot table on the M5 Max"},{"date":"2026-10-05","text":"Mixer x4 and the cache growth rule: the class v3 dataset construction, with the x8 candidate","short":"Mixer x4 and the cache growth rule"},{"date":"2026-10-05","text":"The SP1 CPU prover on PC 1 beside the miners, and the backend survey: no zkVM proves on AMD","short":"The SP1 CPU prover on PC 1 beside the miners, and the backend survey"},{"date":"2026-10-05","text":"Proving v1: segment records, the chain rule, the unproven rule; what was measured tonight","short":"Proving v1: segment records, the chain rule"},{"date":"2026-10-05","text":"Dp4a-class throughput on the M5 Max: the dot4 emulation against the ALU chain","short":"Dp4a-class throughput on the M5 Max"},{"date":"2026-10-05","text":"Layer 3 scratch soundness","short":"Layer 3 scratch soundness"},{"date":"2026-10-05","text":"Epoch length as an era parameter : the Apple M5 Max's compile-ahead per program","short":"Epoch length as an era parameter"},{"date":"2026-10-05","text":"Aggregation cost on the RTX 5090: what a per-block aggregation spends and what each lever gives","short":"Aggregation cost on the RTX 5090"},{"date":"2026-10-05","text":"The C4 fix: certificate-driven reorg","short":"The C4 fix: certificate-driven reorg"},{"date":"2026-10-05","text":"EVM transaction relay: three nodes in a chain, every transaction sent to one end included by the other two miners","short":"EVM transaction relay"},{"date":"2026-10-05","text":"The prover carries both fee tables and the height switch: one pinned guest on either side of DAA 210,000","short":"The prover carries both fee tables and the height switch"},{"date":"2026-10-05","text":"FUD ledger sweep round 6","short":"FUD ledger sweep round 6"},{"date":"2026-10-05","text":"Live devnet: real transactions, the first non-empty shard proven and paid, and the exporter's block structure fixed","short":"Live devnet: real transactions, the first non-empty shard proven and…"},{"date":"2026-10-05","text":"The program id split: why the Apple M5 Max rejected PC 2's proofs, and the verifier at 114 s","short":"The program id split"},{"date":"2026-10-05","text":"Live devnet: the first shards proven, verified and paid","short":"Live devnet: the first shards proven, verified and paid"},{"date":"2026-10-04","text":"One-click Windows workers: what the Apple M5 Max could measure","short":"One-click Windows workers"},{"date":"2026-10-04","text":"The gfx1036 worker fault and what the Apple M5 Max could and could not reproduce","short":"The gfx1036 worker fault and what the Apple M5 Max could and could…"},{"date":"2026-10-04","text":"Difficulty rule v2: the live oscillation, its cause, the DAG replay, the fix behind a height switch","short":"Difficulty rule v2"},{"date":"2026-10-04","text":"The observer stored nothing for 78 minutes, then 7,022 blocks in two minutes","short":"The observer stored nothing for 78 minutes, then 7,022 blocks in two…"},{"date":"2026-10-04","text":"PC 2 at the 14:20 boundary: a worker stuck on the previous epoch","short":"PC 2 at the 14:20 boundary: a worker stuck on the previous epoch"},{"date":"2026-10-04","text":"Shard proving on the RTX 5090: a full shard compressed in 10.9 s, a two-shard block aggregated in 2.2 s, all verified","short":"Shard layer on the RTX 5090: a full shard proven in 10.9 s, a block aggregated in 2.2 s"},{"date":"2026-10-04","text":"First outside machine on the devnet: an Apple silicon laptop through the Igneum Miner app","short":"First outside machine: a friend's Mac mining through the app"},{"date":"2026-10-04","text":"Cloud devnet: 12 igneumd nodes in 5 locations, inter-region latency, two Singapore partitions, hash-rate steps","short":"Cloud devnet: 12 igneumd nodes in 5 locations, inter-region latency"},{"date":"2026-10-04","text":"Correction: the 78-minute observer gap was the Apple M5 Max hibernating","short":"Correction: the 78-minute observer gap was the Apple M5 Max hibernating"},{"date":"2026-10-04","text":"Difficulty v2 rollout rehearsal on the 12-node the cloud provider network, binaries for the devnet, the devnet plan","short":"Difficulty v2 rollout rehearsal on the 12-node the cloud provider…"},{"date":"2026-10-04","text":"Difficulty rule v2 activated on the live devnet at DAA 33,000 by height switch, no fresh chain","short":"Difficulty rule v2 switched on the live devnet by height, no fresh chain"},{"date":"2026-10-04","text":"First machine in the United States: a Windows laptop on an Intel integrated GPU, synced and voting","short":"First machine in the United States: a laptop on an integrated GPU, synced and voting"}]}</script>
 <!-- journey:end -->
 <script>
 (function(){
diff --git a/site/miner.html b/site/miner.html
index ee43b99b..5b166bdd 100644
--- a/site/miner.html
+++ b/site/miner.html
@@ -4,13 +4,13 @@
 <meta charset="utf-8">
 <meta name="viewport" content="width=device-width, initial-scale=1, viewport-fit=cover">
 <title>The Igneum miner app</title>
-<meta name="description" content="One click: install, start, the card mines, and every NVIDIA card from 8 GB proves. Four pages: Mine, Earnings, Prove, Settings. Every card is one row with its rate, watts and pounds a day at your electricity price. Ember Tune measured on an RTX 5090: 84 W saved for 0.15% of rate. A visible 1% software fee you can switch off.">
+<meta name="description" content="One click: install, start, the card mines; a 24 GB NVIDIA card proves as well, measured on an RTX 5090. Four pages: Mine, Earnings, Prove, Settings. Every card is one row with its rate, watts and pounds a day at your electricity price. Ember Tune measured on an RTX 5090: 84 W saved for 0.15% of rate. A visible 1% software fee you can switch off.">
 <link rel="canonical" href="https://igneum.network/miner">
 <meta name="theme-color" content="#0C0C0E">
 <meta property="og:type" content="website">
 <meta property="og:site_name" content="Igneum">
 <meta property="og:title" content="The Igneum miner app">
-<meta property="og:description" content="One click: install, start, the card mines, and every NVIDIA card from 8 GB proves. Four pages: Mine, Earnings, Prove, Settings. Every card is one row with its rate, watts and pounds a day at your electricity price. Ember Tune measured on an RTX 5090: 84 W saved for 0.15% of rate. A visible 1% software fee you can switch off.">
+<meta property="og:description" content="One click: install, start, the card mines; a 24 GB NVIDIA card proves as well, measured on an RTX 5090. Four pages: Mine, Earnings, Prove, Settings. Every card is one row with its rate, watts and pounds a day at your electricity price. Ember Tune measured on an RTX 5090: 84 W saved for 0.15% of rate. A visible 1% software fee you can switch off.">
 <meta property="og:url" content="https://igneum.network/miner">
 <meta property="og:image" content="https://igneum.network/og.png?v=3">
 <meta property="og:image:width" content="1200">
@@ -18,7 +18,7 @@
 <meta property="og:image:alt" content="Igneum. Mined by GPUs. Proven by fire.">
 <meta name="twitter:card" content="summary_large_image">
 <meta name="twitter:title" content="The Igneum miner app">
-<meta name="twitter:description" content="One click: install, start, the card mines, and every NVIDIA card from 8 GB proves. Four pages: Mine, Earnings, Prove, Settings. Every card is one row: rate, watts, pounds a day at your electricity price, a Tune button. A visible 1% software fee you can switch off.">
+<meta name="twitter:description" content="One click: install, start, the card mines; a 24 GB NVIDIA card proves as well, measured on an RTX 5090. Four pages: Mine, Earnings, Prove, Settings. Every card is one row: rate, watts, pounds a day at your electricity price, a Tune button. A visible 1% software fee you can switch off.">
 <meta name="twitter:image" content="https://igneum.network/og.png?v=3">
 <meta name="twitter:image:alt" content="Igneum. Mined by GPUs. Proven by fire.">
 <link rel="icon" href="data:image/svg+xml,%3Csvg xmlns='http://www.w3.org/2000/svg' viewBox='0 0 1024 1024'%3E%3Crect width='1024' height='1024' fill='%230C0C0E'/%3E%3Cg transform='translate(166.95 166.95) scale(6.901)'%3E%3Cpolygon points='50,4 74,34 67,58 80,54 61,96 39,96 20,54 33,58 26,34' fill='%23F2541B'/%3E%3Cpolygon points='50,42 59,58 50,82 41,58' fill='%230C0C0E'/%3E%3C/g%3E%3C/svg%3E" type="image/svg+xml">
@@ -261,8 +261,8 @@ pre b{color:var(--molten);font-weight:500}
   <div class="wrap">
     <div class="hero-copy">
       <div class="eyebrow ember"><span data-product="full">Igneum Ember</span> · one click</div>
-      <h1>Install. Start. The card mines. Every NVIDIA card from 8 GB proves; 12 GB and up mine and prove.</h1>
-      <p class="lead"><span data-product="name">Ember</span> finds your GPU, makes a wallet for you and runs the node, the miner and the prover as one app. Four pages: Mine, Earnings, Prove, Settings. Every card is one row: its rate, its watts, pounds a day at your electricity price, a Tune button. Every hour it compiles the next program while the current one mines, so the card never stops.</p>
+      <h1>Install. Start. The card mines and proves.</h1>
+      <p class="lead"><span data-product="name">Ember</span> finds your GPU, makes a wallet for you and runs the node, the miner and the prover as one app. Every card mines. On this build a 32 GB NVIDIA card mines and proves and a 24 GB card proves a full shard alone, measured on an RTX 5090; 16 GB proves empty shards only; 12 GB and under prove nothing yet (<a href="/evidence#claims" style="color:var(--ember)">evidence, row 16</a>). Four pages: Mine, Earnings, Prove, Settings. Every card is one row: its rate, its watts, pounds a day at your electricity price, a Tune button. Every hour it compiles the next program while the current one mines, so the card never stops.</p>
       <div class="cta">
         <a href="#get" class="btn primary">Downloads: public testnet</a>
         <a href="#features" class="btn">Every feature</a>
@@ -379,7 +379,7 @@ pre b{color:var(--molten);font-weight:500}
     <div class="group reveal" id="start">
       <div class="g-head"><div class="eyebrow ember">01 · One click</div><h3>Install, press Start</h3><p>Five steps from download to the first block on a friend's laptop: drag to Applications, open, Get started, Make me an address, Start mining.</p></div>
       <div class="feats">
-        <div class="feat"><b>The card mines; an NVIDIA card proves</b><p>One button starts the node, waits for sync, starts one miner per card and, on an NVIDIA card, the prover: every NVIDIA card from 8 GB proves; 12 GB and up mine and prove; 24 GB on the stock server (eleven rented cards, RTX 3060 to RTX 5090, 6 October 2026, docs/analysis/prover-tiers-real-cards.md); the patched server for the smaller cards is not yet in the shipped app. AMD and Apple cards mine; a prover for them lands when a zkVM ships one. Quit stops them in order.</p></div>
+        <div class="feat"><b>The card mines; a big NVIDIA card proves</b><p>One button starts the node, waits for sync, starts one miner per card and, on an NVIDIA card, the prover. What proves, measured on an RTX 5090 with the shipped SP1 server (its floor is 13.9 GB): a 32 GB card mines and proves; a 24 GB card proves the full shard alone (from the 5090's allocation, not yet run on a 24 GB card); a 16 GB card proves empty shards only; 12 GB and under prove nothing on this build. Eleven rented cards, RTX 3060 to RTX 5090, were measured on 6 October 2026; the record is pending. AMD and Apple cards mine; a prover for them lands when a zkVM ships one. Quit stops them in order.</p><a class="src" href="/evidence#claims">evidence · row 16</a></div>
         <div class="feat"><b>Finds your GPU by itself</b><p>NVIDIA through the driver, AMD and Intel through OpenCL, Apple silicon through Metal. Real card names, one worker per card.</p></div>
         <div class="feat"><b>A wallet made for you</b><p>A payout key made on your machine and locked to your user, with a backup step before mining starts. Or paste an address you already hold.</p></div>
         <div class="feat"><b>Earnings, money first</b><p>Every block this machine finds pays one address in IGN. The Earnings page counts blocks this run and for life, IGN from proofs, and the electricity cost at your price. £ earned stays 0.00 on devnet and says why; a market price fills it in when one exists (coming). The <a href="/wallet" style="color:var(--ember)">wallet</a> shows the balance.</p></div>
@@ -492,7 +492,7 @@ pre b{color:var(--molten);font-weight:500}
         <tr><td><b>1 · Race the compiler every hour</b></td><td>For each new program, five to ten kernel variants are compiled, benched for two seconds each, and the winner is kept for the hour.</td><td>Measured on the Mac's Metal worker: the winning variant +17.3% on the genesis seed and +21.2% on the hourly seed over the base compile, under load from other work. The RTX 5090 race is prepared and not yet run. <a href="/bench#4-october-2026-miner-performance-variant-racing-metal-worker-on-the-m5-max-the-rtx-5090-job-is-ready-not-run">Log, 4 Oct 2026</a></td></tr>
         <tr><td><b>2 · Auto-tune that learns from the fleet</b></td><td>Apps report the race per card and program class, and every finished Ember Tune. The best settings come back down through the signed update manifest as defaults, so the miner gets faster and more efficient for everyone as the fleet grows.</td><td>Shipped. The fleet is still small, so no fleet table yet; the priors table on <a href="/miners#priors">/miners</a> fills as machines report.</td></tr>
         <tr><td><b>3 · Hash per watt, not hash</b></td><td>Ember Tune: two knobs per card (power limit, core clock), the memory clock held, the point with the best MH per watt within 1% of the top rate kept and pinned. Miners pay for electricity; that is the number they compare.</td><td>Measured on PC 1, 6 Oct 2026: an RTX 5090 from 311.0 W to 226.8 W for 0.15% of its rate (0.411 to 0.563 MH/W); an RTX 4070 from 106.0 W to 75.6 W for none (0.271 to 0.381 MH/W). The chosen clocks are the ladder's floor, not the optimum. AMD through the app's own helper, no prompt; Apple silicon measures only. The <a href="#tuning">table above</a>.</td></tr>
-        <tr><td><b>4 · Template latency</b></td><td>Solo against the local node, a new template within 50 ms of a new tip, because a late block on a BlockDAG goes red and earns nothing.</td><td>Approximate, from the 0.3.6 release plan and not yet in the engineering log: switched p50 46 to 52 ms on a three-node CPU run, 5 Oct 2026. Ships in 0.3.6.</td></tr>
+        <tr><td><b>4 · Template latency</b></td><td>Solo against the local node, a new template within 50 ms of a new tip, because a late block on a BlockDAG goes red and earns nothing.</td><td>Shipped in 0.3.6 (the miner-latency merge; the app is at 0.3.14). Approximate, from the 0.3.6 release plan and not yet in the engineering log: switched p50 46 to 52 ms on a three-node CPU run, 5 Oct 2026.</td></tr>
         <tr><td><b>5 · Never lose a second</b></td><td>Zero-loss hourly program swaps, fault guards, automatic restart, a CPU re-check of every found hash, per-worker health.</td><td>Shipped. Swap 0.01 ms on Metal and 0.00 ms on CUDA with 0 rejected blocks; recoveries in the <a href="#reliability">table above</a>. <a href="/bench#4-october-2026-first-hourly-program-swap-on-the-live-devnet-compile-ahead-no-pause-two-cards">Log, 4 Oct 2026</a></td></tr>
         <tr><td><b>6 · Prove it in public</b></td><td>The bench table per card, fed from the job channel. We claim fastest only when the table says so.</td><td>Live at <a href="/miners">/miners</a>, one row per card, generator version and miner version, each with its log entry.</td></tr>
       </tbody>