System and computer program for efficient cell failure rate estimation in cell arrays

Abstract

A system and computer program for efficient cell failure rate estimation in cell arrays provides an efficient mechanism for raising the performance of memory arrays beyond present levels/yields. An initial search is performed across cell circuit parameters to determine failures with respect to a set of performance variables. For a single failure region the initial search can be a uniform sampling of the parameter space and when enough failure points have been accumulated, a mean is chosen from the mean of the detected failure points. Mixture importance sampling (MIS) is then performed to efficiently estimate the single failure region. For multiple failure regions, a particular failure region is selected by varying the memory circuit cell parameters along a random set of vectors until failures are detected, thus identifying the boundary of the failure region of interest as the closest failure region. A new mean is chosen for MIS in conformity with the location of the detected boundary.

1 . A workstation computer system comprising a processor for executing program instructions and a memory coupled to said processor for storing program instructions, said program instructions including program instructions for efficiently analyzing a failure rate of a cell in an array of cells, said program instructions comprising program instructions for:
 identifying a failure region of interest corresponding to values of circuit parameters of said cell from a result of a statistical performance analysis, wherein the failure region of interest is determined according to first samples of a performance variable;   selecting a sampling function center in conformity with a position of said failure region of interest to increase a frequency of samples of said performance variable in or near said region of interest;   performing a mixture importance analysis in conformity with said sampling function center to determine effects of variation of said cell circuit parameters within and near said failure region of interest on said failure rate, wherein the mixture importance analysis computes second samples from a mixture distribution obtained by summing multiple differing distributions of said performance variable according to weighting coefficients associated with each of said multiple distributions, whereby said second samples are generated according to a mixture of said multiple distributions; and   storing a result of said mixture importance analysis within a memory of said workstation computer system.   
   
   
       2 . The workstation computer system of  claim 1 , wherein said program instructions for identifying include program instructions for:
 uniformly sampling a cell circuit parameter space to select input vectors for said performance analysis; and   accumulating at least a threshold number of failure points to define said failure region of interest.   
   
   
       3 . The workstation computer system of  claim 2 , wherein said program instructions for selecting compute a center of gravity of said failure region of interest from input vectors of said accumulated failure points. 
   
   
       4 . The workstation computer system of  claim 2 , wherein said program instructions for identifying comprise program instructions for:
 varying said cell circuit parameters from a nominal-valued vector of said cell circuit parameters until a failure point is detected over a number of randomly generated cell circuit parameter vector directions; and   finding a group of failure points resulting from said varying and closest to said nominal-valued vector to locate a boundary of said failure region of interest, whereby said failure region of interest is identified.   
   
   
       5 . The workstation computer system of  claim 4 , wherein said program instructions for selecting compute a mean input vector across said group of failure points, whereby said new sampling center is selected around said boundary of said failure region of interest. 
   
   
       6 . The workstation computer system of  claim 4 , wherein said program instructions for selecting estimate a point within said failure region of interest from input vectors corresponding to points within said group of failure points, whereby said new sampling center is selected within said failure region of interest. 
   
   
       7 . The workstation computer system of  claim 6 , wherein said estimating estimates the center of said failure region of interest. 
   
   
       8 . The workstation computer system of  claim 2 , wherein said program instructions for performing said mixture importance analysis are executed to determine each of a plurality of performance variables including writeability, read stability, write delay and read delay times. 
   
   
       9 . The workstation computer system of  claim 1 , wherein said multiple distributions comprise at least three distributions, and wherein said program instructions for performing perform said mixture importance analysis in conformity with said at least three distributions. 
   
   
       10 . A computer program product comprising a computer-readable storage media encoding program instructions for execution on a workstation computer, said program instructions for altering design parameters of a cell in an array of cells, said program instructions comprising program instructions for:
 identifying a failure region of interest corresponding to values of circuit parameters of said cell from a result of a statistical performance analysis, wherein the failure region of interest is determined according to first samples of a performance variable;   selecting a sampling function center in conformity with a position of said failure region of interest to increase a frequency of samples of said performance variable in or near said region of interest;   performing a mixture importance analysis in conformity with said sampling function center to determine effects of variation of said cell circuit parameters within and near said failure region of interest on said failure rate, wherein the mixture importance analysis computes second samples from a mixture distribution obtained by summing multiple differing distributions of said performance variable according to weighting coefficients associated with each of said multiple distributions, whereby said second samples are generated according to a mixture of said multiple distributions; and   storing a result of said mixture importance analysis within a memory of said workstation computer system.   
   
   
       11 . The computer program product of  claim 10 , wherein said program instructions for identifying include program instructions for:
 uniformly sampling a cell circuit parameter space to select input vectors for said performance analysis; and   accumulating at least a threshold number of failure points to define said failure region of interest.   
   
   
       12 . The computer program product of  claim 11 , wherein said program instructions for selecting compute a center of gravity of said failure region of interest from input vectors of said accumulated failure points. 
   
   
       13 . The computer program product of  claim 11 , wherein said program instructions for identifying comprise program instructions for:
 varying said cell circuit parameters from a nominal-valued vector of said cell circuit parameters until a failure point is detected over a number of randomly generated cell circuit parameter vector directions;   finding a group of failure points resulting from said varying and closest to said nominal-valued vector to locate a boundary of said failure region of interest, whereby said failure region of interest is identified.   
   
   
       14 . The computer program product of  claim 10 , wherein said multiple distributions comprise at least three distributions, and wherein said program instructions for performing perform said mixture importance analysis in conformity with said at least three distributions.