igneum-network/igneum
1
0
Fork 0
Code Issues Pull requests Releases Activity
igneum/docs/analysis
History
igneum-labs 9417a14ac7 Counter ASIC 3.0 item 8: the RTX 5090 rows, the chip side on the measured 11 pJ per op, the verdict 
PC 2 job run-ca3-shadow-pc2-20261006 (card empty, every pack bit-exact against the Mac): the 5090 holds its rate to
150,800 ops per hash and loses 2.7 percent at 199,600 under the app's 431 W cap, which binds from 102,100 ops up and
takes the clock from 3,037 to 1,834 MHz (86 MH/s at 330,700 ops); 350 W at the control, 2.65 to 3.27 microjoules per
hash; marginal ALU energy 10 to 13 pJ per counted op. Clock rows OWED (nvidia-smi refused -lgc without rights). Chip
side at N = 100,000 and k = 1: 2.1x over the 5090 on GDDR7, 0.9x over the M5 Max. GO at mx8+sh256x27.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-10-06 08:44:57 +00:00
..
amd-proving.md docs/analysis/amd-proving.md: no zkVM proves on AMD (SP1, RISC Zero, Jolt, OpenVM, ICICLE cited), the SP1 CPU prover measured on PC 1 beside the miners (282 s a shard at any size, 30 GB RSS: no CPU tier), the tier consequences and the public line; PC 1 job scripts with the bash -n gate; bench-log entry 2026-10-05 21:02:59 +00:00
asic-resistance-history.md docs/analysis/asic-resistance-history.md: the 2011 to 2026 history of ASIC resistance (31 rows, chip, gain per joule, months held, response), the papers, the chip economics and the market-cap table, ten lessons, the audit of class v2 and the Counter ASIC 2.0 layers, seven ranked additions (partial-store chip row first), nine candidates placed nowhere 2026-10-05 22:38:17 +00:00
base-fee-floor.md Prover mirrors the fee switch: both tables and fees_v1_activation_daa in the shard input, guest re-pinned, devnet runbook for H = 210,000 2026-10-05 16:24:35 +00:00
card-lifetime-2026-10-05.md CI on the merged 0.3.11 tree: MacBook reworded in the analysis prose (the identity check's hostname pattern), a presence check before the kit's first use in the three proving-v1 PC 2 scripts (C32), pc1-cpu-prove.ps1 on the socket check's allow list (the CPU path starts no GPU server) 2026-10-05 22:44:09 +00:00
chip-model-v3.md Counter ASIC 3.0 item 1: the partial-store chip and the time-memory curve in chip-model-v3.md section 5 2026-10-06 07:32:00 +00:00
difficulty-2026-10-03.md Difficulty: timestamp attack fixed in spec 2.3, simulator, analysis addendum, ledger M23 2026-10-04 00:56:20 +00:00
difficulty-2026-10-04-oscillation.md Difficulty: the live oscillation of 4 October, its cause, the DAG replay and rule v2 (reference window 600 behind a height switch) 2026-10-04 13:26:23 +00:00
economy-2026-10-04.md Economy: agent-based mining-versus-proving simulation, six stress scenarios, lever study, analysis and bench entry 2026-10-03 23:16:46 +00:00
hashrate-decay-2026-10-03.md Hash-rate decay on the RTX 5090: diagnosis, Metal reproduction, proposed Seeder fix 2026-10-03 23:31:23 +00:00
hashrate-decay-2026-10-03.patch Hash-rate decay on the RTX 5090: diagnosis, Metal reproduction, proposed Seeder fix 2026-10-03 23:31:23 +00:00
int8-matrix-family.md Counter ASIC 2.0 layer 7: RTX 5090 and RX 9070 XT dot4 numbers from PC 1 (job run-dot4-20261005): dp4a 7.45 T/s on the 5090 via inline PTX, v_dot4_i32_iu8 0.66 T/s on the 9070 XT via the clang builtin in Adrenalin OpenCL C, signed emulation 7.1x / 1.46x / 4.7x the ALU step on NVIDIA / AMD / Apple; no PC platform lists cl_khr_integer_dot_product 2026-10-05 20:31:33 +00:00
latency-shadow-2026-10-06.md Counter ASIC 3.0 item 8: the RTX 5090 rows, the chip side on the measured 11 pJ per op, the verdict 2026-10-06 08:44:57 +00:00
m16-recompute-attacker-2026-10-05.md FUD sweep round 6 (evening, 5 October): eleven rolled-out fixes moved to Fixed with live measurement lines, F21/F22 shipped but switch not thrown, M20 closed on the 0.3.5 merge; P14 one base-fee definition in spec 05; M16 recompute-attacker cost model; C4 overlay-against-GHOSTDAG runner 2026-10-05 16:02:02 +00:00
prover-floor.md Prover floor: sweep 1 table and reading (the Setup-time keys bind at 12.7 GB) 2026-10-05 22:42:17 +00:00
proving-methods.md CI on the merged 0.3.11 tree: MacBook reworded in the analysis prose (the identity check's hostname pattern), a presence check before the kit's first use in the three proving-v1 PC 2 scripts (C32), pc1-cpu-prove.ps1 on the socket check's allow list (the CPU path starts no GPU server) 2026-10-05 22:44:09 +00:00
scratch-soundness.md ca2-soundness: docs/analysis/scratch-soundness.md alone (its code is already in the tree through the mixer merge; the branch merge conflicted there); release-0.3.11 plan: C37 and C38 2026-10-05 23:04:38 +00:00
security-budget.md Evidence page, end-to-end proving standard, funding plan, security-budget stress, payment routes 2026-10-03 22:35:03 +00:00
sram-mirror.md Counter ASIC 2.0 layer 6, revision 2: shipped cache-die density (AMD V-Cache 64 MB on 41 mm^2 at 7 nm, Hot Chips 33) as the headline column beside the bit-cell lower bound; mirror 106 to 164 mm^2 and $30 to $56 per die at 256 MiB; year 0 to 10, hot-table and option tables redone; latency citations section (MEMSYS 2018, Chang 2017, mining-chip memory types) 2026-10-05 20:20:46 +00:00
weak-program-census-2026-10-03.md Mac app: dial the public seed node first, the home node second 2026-10-03 22:28:58 +00:00