diamondBuilt for Corning

The Glass TGV Impedance Design Kit

Calibrated to Eagle XG. 3.57% MAE against 5 published IEEE papers. Your customers are designing blind without this data.

The coaxial closed-form approximation that every other tool relies on is wrong by 55–225% on glass substrates. We verified this across 100 randomly sampled geometries.

Request Eagle XG DemoOpen Buyer App
3.57%MAE vs IEEE
2.95MBEM Simulation Rows
16Glass Types
995Curated S2P Assets
12Named Golden Configs
909Filed Sub-Claims280 utility + 629 provisional · 146 indep · 0 issued · deadline 2027-01-31
The Problem

Your Customers Are Designing Blind on Glass

Every RF engineer knows the coaxial formula. But on glass substrates with through-glass vias, that formula produces 55–225% error. No existing EDA tool ships a glass-calibrated impedance model. Corning sells the substrate—but without a design kit, adoption stalls.

report_problem

Coaxial Formula Fails on Glass

55–225% impedance error on TGV geometries. Designers cannot trust the numbers for Eagle XG routing.

check_circle

BEM Solver Calibrated to IEEE

3.57% MAE against Sukumaran ECTC 2014, Watanabe ECTC 2019, Shorey JMS 2016, Tummala JEP 2020, Hwang TMTT 2017.

Eagle XG Design Coverage

memoryUCIe @ 32 GT/s
storageDDR5 @ 6.4 Gbps
developer_boardPCIe Gen6
lan400G Ethernet
wifiWiFi 7
radarAutomotive Radar
grid_viewDense I/O
radioLow-Loss RF

Licensing Partnership

inventory_2

Bundle with Eagle XG

Ship the Genesis Glass PDK alongside Eagle XG substrate sales. Your customers get a design kit that actually works for glass. You sell more substrates.

shield

909 Filed Sub-Claims

Comprehensive IP across BEM impedance, yield prediction, isolation synthesis, bondability screening, and provenance-bound signoff workflows. 280 utility filed Jan 2026 + 629 provisional (146 independent). Provisionals establish priority date but are not examined or enforceable; non-provisional conversion deadline 2027-01-31; 0 claims issued to date. Full claim text and prosecution status under NDA.

speed

Sub-100ms Results

BEM impedance results in under 100 milliseconds via live API. Coaxial approximation delta shown side-by-side. The current buyer surface exposes 261 `/v1` routes, a 25-case package-signoff benchmark, and 982 repo-root tests collected.

Platform at a Glance

Validated Solvers32+ Production
BEM Database15.92M Rows + 901K Versioned Strict Additions
S2P Touchstone Assets995 active curated files
Glass Types16 Substrates
Metals Supported5 Types
Frequency Bands10 Bands
Tests Collected982
Patent Claims909 Filed
Dk 5.27

Eagle XG at 10 GHz — Primary Reference Substrate

55–225%

Coaxial Formula Error on Glass — Verified 100 Geometries

NDA

Replication Cost & Valuation — Under NDA Data Room

Give Eagle XG Customers the Design Kit They Need

15-minute screen-share demo. Run the impedance calculator on Eagle XG parameters live. NDA + data room access within 24 hours.

Schedule 15-Min DemoOpen Buyer App
Interface signoff
Eagle XG · UCIe / HBM4 screening →
Show your customers what UCIe-on-Eagle-XG and HBM4-on-Eagle-XG actually budget out to, before they buy a panel.
literature provenance
Source-typed witness corpus →
Eagle XG appears in 2 of the 5 published Z₀ references in our corpus. Both are tagged source_type:simulation.
Productization fleet
5 net-new endpoints — your customers can RFQ a coupon today
POST /v1/coupons/export-fab — Eagle XG / Iris / MEA stack-up + Amkor SOW + $55K-$165K cost band
POST /v1/glass-pdk/geometry-pareto — multi-objective design across Eagle XG dispersion
POST /v1/glass-pdk/drc-validate — manufacturability + plating clearance + aspect ratio rules
POST /v1/glass-pdk/validate-against-measurement — surrogate-vs-literature with honest sim-vs-VNA flag (closes the "where's the VNA" question)
GET /v1/glass-pdk/cross-solver-matrix — proves BEM agreement with Palace FEM on 100 geometries
Coupon shop integration: Amkor / MOSIS / generic templates. Stack-up generator includes Eagle XG-specific glass thickness defaults.
Per-regime cross-physics cal heads · v2-clean retrain (2026-05-05)
6/6 regimes deployed · ULTRA_HIGH_FREQ 90.5% · HBM4 90.3%
First successful cross-physics calibration on Genesis. Per-regime BEM-vs-Palace μ-correction across all 6 glass-substrate regimes — closes BEM-quasi-static-vs-Palace-full-wave physics envelope. Pooled 17.10% → 4.71% (72.5% pooled gap closure, avg 74.0% across regimes, 4/6 elite >80%): ULTRA_HIGH_FREQ 90.5% · HBM4 90.3% · WIDE_PITCH 87.6% · UCIE 81.6% · EXTREME_TIGHT 50.6% · MMWAVE 43.4% gap closure. HBM4 head trained on 4507 Palace runs @ 50 GHz on Eagle XG-class glass (d ∈ [25,50] µm, p ∈ [40,80] µm, t ∈ [300,600] µm). Pattern reproduces across 6/6 regimes — not a single-experiment artifact.
Track A · DD-grade coverage
95% conformal interval + per-axis OOD on every prediction
Every /v1/glass-pdk/predict-impedance response carries a distribution-free 95% Palace-truth coverage interval — HBM4 q=8.38 Ω, cov 94.46%; UCIE q=6.03 Ω, cov 95.41%; EXTREME_TIGHT q=35.10 Ω, cov 94.42%; WIDE_PITCH q=9.84 Ω, cov 94.44%; pooled fallback q=9.11 Ω — AND a per-axis OOD severity diagnostic across all 7 input axes. For a glass substrate vendor, this means every Z₀ claim Corning attaches to an Eagle XG (or Iris / MEA) datasheet has a defensible mathematical bound, not a marketing approximation.
Track B · Eagle XG visibility via OSS distribution
pip install glass-tgv-diffpair · Eagle XG in the headline universal-law table
Apache-2.0 OSS slice pip install glass-tgv-diffpair publishes the universal scaling law log(Z₀_diff)=0.338·log(sep/d), R²=0.918, n=2.1M — Eagle XG appears in the headline 5-glass table alongside AF32, Borofloat33, EN-A1, Fused Silica with exponent 0.3385 (vs universal 0.338, drift 0.001 / 0.3%). Every academic citation of the OSS package is an unpaid impression for the Eagle XG brand. Companion DAC 2026 paper drafted. 30 MCP tools + 10 packaged Claude Desktop / Cursor / Codex agents (HBM4 Signoff / Inverse Design / DRC Fixer / Pareto Explorer / + 6 more) wire this into AI-agent workflows.
Composite lab_readiness_score
One 0-100 score on every Eagle XG impedance response
Single 0-100 score with verdict bands (send_to_lab ≥95 / send_with_extra_qc 80-94 / hold 60-79 / reject <60) on /predict-impedance + /geometry-from-target + /coupons/export-fab. Pro-rata weights across 5 active confidence gates (cross_solver 24 + conformal 24 + per_axis_ood 18 + public_data 18 + ensemble 16 = 100). For Corning customers selecting between Eagle XG geometries, the score gates whether the BEM Z₀ claim is fab-ready or whether more cross-physics evidence is needed before shipping a coupon design. No fabricated proxy values: any gate without measured signals contributes None and triggers partial_score=true; fab-coupon export refuses bundles when partial.
3/5 cross-solver matrix wired
BEM ✓ Palace ✓ FastHenry2 ✓ · OpenEMS / gprMax honest skip
100-geometry × N-solver disagreement matrix at 3 of 5 wired (+ 4/5 spot-checks at 1-geom and 10-geom; witness benchmarks/sprint42/cross_solver_matrix_{1geom,10geom}_4of5_2026_05_01.json). BEM always-available + Palace via spack-installed binary + FastHenry2 via local fasthenry binary (Z₀ from L_FH + analytical cylindrical-coax C — hybrid honestly disclosed; dispatch key = FastHenry2_L_plus_analytical_C). OpenEMS + gprMax return None via try-import skip when binary not on PATH. Witness: docs/HONEST_FALLBACKS.md Fallback 5.