8 casesBounded case budgetsCOMPLETEInspect metric contract
4 classesExplicit uncertainty classesCOMPLETEInspect metric contract
26 fixturesDeterministic contract fixturesCOMPLETEInspect metric contract
SIMULATION / SA-MET-VARIABILITY-UQ-068
Variability, metrology & UQ workbench
Separate uncertainty before combining it.
Move from LER/LWR bandwidth and finite sampling to hierarchical SRAM variation, plasma-etch uncertainty, Gauge R&R, and hybrid-bonding metrology without mixing quantity, grain, denominator, or evidence class.
EXPLORE
Start with the observable.
Use this page when the simulation is the primary artifact. The linked lesson explains the concept; the linked lab keeps the visual replay surface available.
- Select one case and inspect aleatoric, epistemic, measurement, and model contributions without treating Type A or Type B as uncertainty classes.
- Move correlation length, aperture, and sample count to see how extraction bandwidth, spatial filtering, and finite degrees of freedom change interpretation.
- Inspect hierarchical variance closure and refusal states before exporting a same-contract teaching budget.
V136 / CURRENT-TAB WORKBENCH
Keep the contract attached to the number.
Explore one bounded case at a time. Quantity, unit, grain, denominator, and evidence class remain fixed while local teaching assumptions move.
8 case budgets4 uncertainty classes0 network writes
Patterningedge-position residualdetrended edge-trace segment / per 256 nm trace
Combined standard uncertainty0.4379 nm
Same quantity, grain, denominator, and illustrative evidence class only.
k = 2 teaching interval+/- 0.8759 nm
Not a universal confidence, coverage, acceptance, or disposition statement.
Effective degrees of freedom122.4
Welch-Satterthwaite teaching calculation over finite component degrees.
componentrelative variance
ALEATORICTYPE A
EPISTEMICTYPE B
MEASUREMENTTYPE A
MODELTYPE B
Covariance contribution: 0.0054 nm2
Quantityedge-position residualUnitnmGraindetrended edge-trace segmentDenominatorper 256 nm traceEvidenceillustrative-original-surrogate
RUN CARD
Conditions before conclusions.
The browser evaluates original teaching fixtures only. It does not fit measurements, operate equipment, recommend settings, establish acceptance, qualify a process, dispose material, or run a production solver.
Primary knobs
- case-local component scales
- named correlation term
- correlation length
- measurement aperture
- sample count
Saved outputs
- combined standard uncertainty
- k=2 teaching interval
- visible PSD variance fraction
- aperture variance ratio
- finite-sample sigma uncertainty
- hierarchy closure
- deterministic JSON and CSV
EXPORT
Download the public artifacts.
These exports are the supported public files for review. PDK files and restricted third-party material are not redistributed.
V136 public model and fixture
/data/semiagora-variability-metrology-uq-v1.json
Open file ->METHODS
How the number was made.
Methods are written for citation discipline: the extraction rule matters as much as the plotted value.
Budget terms combine only inside one declared quantity, unit, grain, denominator, and evidence class, including explicit named covariance terms.
Keep this method attached when reusing the figure or metric.
A Lorentzian teaching PSD is integrated over the locked measurement-length and Nyquist bounds; aperture filtering uses a sinc-squared boxcar transfer.
Keep this method attached when reusing the figure or metric.
Finite-sample sigma uncertainty uses a first-order normal-independent approximation with n-1 degrees of freedom, while hierarchy closes only over declared fixed-condition variance components.
Keep this method attached when reusing the figure or metric.
LIMITS
What this page does not claim.
The MVP is useful because the limits are visible. These statements prevent a teaching simulation from being cited as silicon evidence.
All values are original deterministic teaching fixtures, not fab, wafer, device, bonding, or production measurements.
Carry this boundary into any derivative note, paper draft, or slide.
The Lorentzian PSD, sinc-squared aperture response, finite-sample approximation, and component budgets are explicit teaching models, not universal physical laws.
Carry this boundary into any derivative note, paper draft, or slide.
Type A and Type B are evaluation methods and do not replace the aleatoric, epistemic, measurement, or model classification.
Carry this boundary into any derivative note, paper draft, or slide.
Values from different quantities, units, grains, denominators, or evidence classes are not numerically ranked.
Carry this boundary into any derivative note, paper draft, or slide.
No process recipe, acceptance threshold, qualification, disposition, signoff, or production action is produced.
Carry this boundary into any derivative note, paper draft, or slide.
EVIDENCE
Trace the claim back to artifacts.
The evidence route remains the catalog-level browser; this page is the simulation-level reading card.
The immutable V136 contract independently verifies 1,210 checks and 26 deterministic acceptance fixtures before implementation.
SA-MET-VARIABILITY-UQ-068
Fifteen link-only source anchors provide terminology and replacement-evidence context; no source body or source numeric coefficient enters the model.
SA-MET-VARIABILITY-UQ-068
Eight refusal states preserve mixed-dimension, missing-class, unsupported-correlation, insufficient-sample, cross-case, and production boundaries.
SA-MET-VARIABILITY-UQ-068
Source spine
These links anchor the public teaching model. They do not convert the replay into a qualified process recipe or signoff result.
NIST: Issues in Line Edge and Linewidth Roughness Metrologyextraction bandwidth and reporting boundaryNIST: Line Edge and Linewidth Roughnessroughness terminology and spectral contextNIST: Dimensional Metrology for Nanoscale Patternsmeasurement contextNIST/SEMATECH: nested variationhierarchical design contextNIST/SEMATECH: variance componentsvariance-component terminologyNIST/SEMATECH: confidence interval for standard deviationfinite-sample uncertainty contextNIST TN 1297: Reporting Uncertaintyuncertainty reporting boundaryBIPM VIM: measurement precisionrepeatability and reproducibility vocabularyBIPM JCGM publicationsGUM and VIM publication contextNIST/SEMATECH: Gauge R&R overviewmeasurement-system decompositionNIST/SEMATECH: uncertainty from a gauge studygauge-study uncertainty boundaryNIST: metrology and process-control measurement uncertaintydecision-risk contextNIST: hybrid-bonding-ready nanoscale mechanical characterizationhybrid-bonding metrology signalSemiAgora V127 LER/LWR variability chainpublic synthetic roughness inputSemiAgora V128 SRAM variation coveragepublic synthetic hierarchy input
Public execution boundary
No server-side live ngspice execution, uploads, accounts, payments, comments, newsletter signup, or job/event submissions are active on this page.