Free Online Cpk / SPC Calculator
Paste your measurement data and instantly generate Cpk, Cp, Ppk analysis with control charts, process capability gauges, sigma level, and yield estimates — completely free, no sign-up required
Specification Limits & Settings
Process Data
Data Validation
Process Capability Overview
Cpk Gauge
Normality Assessment
Descriptive Statistics
| Metric | Value | Metric | Value |
|---|
Control Charts
X-bar Control Chart (Individual Values)
Moving Range Chart
Data Distribution
Histogram with Normal Curve Overlay
Still analyzing Cpk manually? Let AI optimize it for you
NeuroBox delivers real-time Virtual Metrology, Run-to-Run auto-tuning, and Equipment Intelligence to keep Cpk consistently above 1.67 — boosting yield by 8%+.
Request a Demo Explore SolutionsFrequently Asked Questions
What is Cpk (Process Capability Index)?
Cpk measures how well a process fits within its specification limits, accounting for process centering. A Cpk of 1.0 means the process just fits within specs (3-sigma), 1.33 is the typical minimum requirement for production (4-sigma equivalent), and 1.67 or above is considered excellent (5-sigma equivalent). Unlike Cp, Cpk penalizes processes that are off-center from the target.
What is the difference between Cp/Cpk and Pp/Ppk?
Cp and Cpk use within-subgroup variation (short-term capability) estimated via the R-bar/d2 method, reflecting inherent process variation. Pp and Ppk use overall (total) variation from all data points, capturing both short-term and long-term variation. When subgroup size is 1, Cp/Cpk are estimated using the Moving Range method (MR-bar/d2 with d2=1.128). If Cpk is significantly higher than Ppk, it indicates the process has shift-to-shift or lot-to-lot variation that inflates overall spread.
How do I use subgroup size?
Subgroup size determines how measurements are grouped for within-subgroup variation estimation. Use subgroup size = 1 for individual measurements (the calculator uses the Moving Range method). Use subgroup size 2-25 when consecutive measurements form natural groups, such as multiple wafers from the same lot, or multiple measurements from the same cassette. The data is split into consecutive subgroups, and within-subgroup range (R-bar) combined with the d2 constant estimates short-term sigma.
What does the Process Sigma Level mean?
Process Sigma Level (also called Z-score or sigma capability) indicates how many standard deviations fit between the process mean and the nearest specification limit. A 3-sigma process has a Cpk of 1.0 and yields ~99.73%. A 6-sigma process (Cpk = 2.0) yields ~99.99966% with only 3.4 DPMO. In semiconductor manufacturing, Cpk ≥ 1.33 (4-sigma) is typically the minimum standard for production release, and Cpk ≥ 1.67 (5-sigma) is required for critical dimensions.
How to use this SPC calculator?
1) Enter your Upper Specification Limit (USL) and Lower Specification Limit (LSL). Optionally set a Target value and subgroup size. 2) Paste your measurement data into the text area — the tool accepts values separated by lines, commas, spaces, or tabs. 3) Click "Run Analysis" to generate a complete process capability report including Cpk/Ppk indices, control charts, histogram with normal distribution overlay, sigma level, yield estimate, and DPMO. You can also click "Load Sample Data" to try with example semiconductor wafer thickness data.
Why does normality matter for Cpk?
Cpk calculations assume the data follows a normal (Gaussian) distribution. If the data is highly skewed or has heavy tails (high kurtosis), the Cpk value may be misleading — the actual defect rate could be higher or lower than predicted. This calculator checks skewness and excess kurtosis: values within ±1.0 generally indicate acceptable normality. If normality is questionable, consider using Ppk with actual defect counts or non-parametric capability methods.
Ready to make process control smarter?
NeuroBox integrates Virtual Metrology (VM), Run-to-Run (R2R) auto-tuning, and Equipment Intelligence Platform (EIP) for fully automated closed-loop process control — boosting yield by 8%+.