Test data analysis for Windows

From raw test data to yield answers, in one desktop application.

Testalytix analyses and visualises semiconductor and measurement test data, with optional Python scripting and a Claude AI assistant. Built for test, yield, product and quality engineers working with wafer-test, final-test and qualification data.

Testalytix overview
Click the image to view it full size.
Testalytix Plot tab showing a parametric wafer map with side histogram, colour scale and statistics table

Reads your data as it comes off the tester

  • STDF (incl. .gz / .zip)
  • CSV / TXT
  • EDF
  • Excel
  • SQL Server

One data set, many tools

Load once. Every analysis works on the same data.

Open several files together and they are merged into one spreadsheet-like grid. Clean it there, and every statistic, chart and report follows.

01 / DATA

Data grid & cleaning

Edit, sort, filter, group, transpose and de-duplicate. Mark and remove outliers, remove failing rows and add calculated columns.

02 / STATISTICS

Statistics tables

Count, mean, median, standard deviation, Cp/Cpk, %Fail and more, for all tests at once or per group such as lot or wafer.

03 / LIMITS

Flexible limits

Load limits from a file, type them in, or calculate them: PAT-based or from a Cpk target.

04 / YIELD

Bin & yield analysis

Fail Pareto, Generate Bin1 and Retest Recovery, with the STDF hardware and software bin records at hand.

05 / DATABASE

SQL Server workspace

Query your test database directly with a query builder and SQL editor, and analyse the result like any file.

06 / COMPARE

Pre / post measurement

Compare rounds before and after stress, burn-in, rework or retest, device by device, to find new fails and parameter shifts.

Charts

Eight chart types, one double-click away.

Pick a chart type, double-click a test, and step through the test list with the arrow keys. Group by lot or wafer, zoom into a region to filter the data, and see the statistics of exactly what is plotted.

  • Box plot
  • Histogram
  • Cumulative P%
  • XY plot
  • Temperature evaluation
  • Wafer map (parametric & bin)
  • Defect density
  • Scatter plot
Box plot of one test grouped by three lots, with limit lines and a per-lot statistics table
PCA analysis window with a biplot of scores and loadings, eigenvalue table and loadings matrix

Advanced analysis

Find what drives the variation.

Go beyond single-test statistics with multivariate and measurement-system tools built into the same workflow.

Correlation & drift
XY correlation between tests and pre/post drift analysis.
PCA & K-Means clustering
Scree, scores, loadings and biplots; cluster profiles to spot device populations.
Gauge R&R / MSA
Separate measurement-system variation from part variation, prober to prober.

Reporting & re-use

Reports in the same window. Setups you can repeat.

The built-in report editor collects charts and statistics tables into a document you can save, export and print. A plot recipe stores a whole chart setup and re-applies it to the next lot in two clicks.

Report editor with a word-processor ribbon and a cumulative plot with its statistics inserted into the page

Report current, selected or by category

One plot, a list of ticked tests, or every test category in one run.

Full word-processor editor

Document tabs, spell checking, and print, draft, web and reading layouts.

Save as HTML

Share results with colleagues who do not have the application.

Plot recipes

Save a chart setup once and apply it to every new lot.

Python & Claude

Your own scripts. An AI assistant that knows your data.

When a calculation needs something no button offers, the Python page runs your own scripts on the loaded data, and the Claude chat helps you write them.

Python on the grid

Read the data grid into pandas, analyse it with numpy, scipy and the rest of the ecosystem, and send the result table back for charts and reports.

Uses your Python

Finds 64-bit Python 3.8+ installations, virtual environments and conda environments on the PC by itself.

Ask Claude about your data

Ask in your own words about your data, a plot or a wafer map. Claude answers, writes the Python script, and, when you ask, plots, reports, sorts and filters.

You stay in control

Runs with your own Anthropic API key, stored encrypted on your PC. The assistant is limited to Testalytix tools; sorting, filtering and several plot actions ask Yes / No first.

Built for long days of engineering work.

Sharp on any screen

Windows, fonts and icons adapt to each monitor's scaling, also when moved between monitors.

Clear feedback

Actions that cannot run explain why, and long operations show progress until they finish.

STDF records included

Test synopsis, master information, hardware and software bin tables alongside your data.

Full user manual

Every ribbon page, chart and analysis documented step by step, with built-in help.

Contact

Interested in Testalytix?

Get in touch for a demonstration, licensing options or any further information. Tell us about your test data and we will show you how Testalytix fits your workflow.