Calculator inputs
Use the responsive grid below for large, medium, and mobile screens. Results appear above this form after submission.
Example data table
| Wafer (mm) | Die Size (mm) | Scribe (µm) | Defect Density | Model | Gross Dies | Good Dies |
|---|---|---|---|---|---|---|
| 300 | 10 × 8 | 80 × 80 | 0.35 /cm² | Negative Binomial | 738 | 522.43 |
| 200 | 7 × 7 | 60 × 60 | 0.45 /cm² | Murphy | 516 | 386.48 |
| 150 | 5 × 5 | 50 × 50 | 0.22 /cm² | Poisson | 574 | 513.24 |
Formula used
1) Effective wafer diameter
Deff = D − 2E, where D is wafer diameter and E is edge exclusion.
2) Gross die area with scribe lanes
Agross = (W + Sx) × (H + Sy), using die width W, die height H, and street widths Sx and Sy.
3) Gross dies per wafer
GDPW ≈ (πDeff² / 4Agross) − (πDeff / √(2Agross)), then scaled by layout utilization.
4) Critical area and defect load
Ac = Aactive × critical area factor and m = D0 × Ac, where D0 is defect density in defects/cm².
5) Yield models
Poisson: Y = e−m
Murphy: Y = ((1 − e−m) / m)²
Negative Binomial: Y = (1 + m/α)−α
6) Final shipped good dies
Good Dies = GDPW × Fab Yield × Parametric × Probe × Final Test × Assembly
How to use this calculator
- Enter wafer diameter and edge exclusion to set the usable wafer area.
- Provide die width, die height, and scribe lanes for gross die footprint.
- Set layout utilization to reflect stepping inefficiency and edge packing losses.
- Enter defect density and critical area factor for defect-sensitive die exposure.
- Choose the yield model that best matches your fab clustering behavior.
- Fill in downstream yields for parametric, probe, final test, and assembly stages.
- Optionally add wafer and backend costs to estimate cost per good die.
- Press Calculate Yield to show the result above the form, then export CSV or PDF if needed.
FAQs
1) What does gross dies per wafer mean?
It is the estimated count of die locations that fit on the usable wafer after considering die footprint, edge exclusion, and layout losses.
2) Why include scribe lane width?
Scribe lanes consume wafer area between adjacent dies. Ignoring them overstates die count and makes yield planning look better than reality.
3) When should I use the Poisson model?
Use Poisson when defects are assumed random and independent across the die. It is simple and usually gives a conservative first-pass yield estimate.
4) Why does Murphy yield differ from Poisson?
Murphy relaxes the fully random assumption and often predicts higher yield at the same defect density, especially for moderate defect loads.
5) What is the clustering factor alpha?
Alpha controls how strongly defects cluster in the negative binomial model. Lower alpha means more clustering and larger differences from Poisson yield.
6) Why are downstream yields separate?
A die can survive random defect screening yet still fail parametric, probe, package, or final test steps. Separate fields show where losses accumulate.
7) Is this suitable for production signoff?
It is best for planning, comparison, and sensitivity analysis. Production signoff should use fab-specific maps, SPC data, and product qualification history.
8) How is cost per good die calculated?
The tool divides wafer cost by shipped good dies, then adds backend cost per die. Fewer good dies increase the effective unit cost quickly.