Analytical model for predicting current mismatch in metal oxide semiconductor arrays

A method of predicting current mismatch in a semiconductor array comprising the steps of:

• providing a semiconductor substrate having a plurality of rows and columns defining a set of cells;
• selecting a first subset of cells of the set;
• measuring current for each cell in the first subset of cells;
• determining a local variation of current for one or more cells in the first subset;
• determining variation of current due to a poly density gradient effect in a first direction based on the local variation and the variation of current for one or more cells in the first subset;
• performing computations in a processor to predict current mismatch in a second subset of cells of the set based on the variation of current due to the poly density gradient effect in the first direction, the second subset comprising one or more cells each mutually exclusive of the first subset of cells.

A computer system comprising:

• a non-transitory computer readable storage medium encoded with computer program code;
• and
• a processor configured to execute the computer program code for characterizing an array, such that when the processor executes the code, the processor performs a method comprising:
• selecting a first subset of cells from a plurality of cells in the array;
• determining variation of measured current for each cell in the first subset of cells with respect to current measured for a reference cell in the array;
• determining local variation of current for one or more cells in the first subset;
• and
• determining variation of current due to a poly density gradient effect in a first direction based on the determined local variation and the determined variation of current for one or more cells in the first subset.