Process Variability Effect on Soft Error Rate by Characterization of Large Number of Samples

For the first time die-to-die soft error rate (SER) variability from process manufacturing is experimentally characterized by irradiating a very large number of samples from a single wafer. Alpha and neutron SER is measured and reported as a function of original location on wafer, test dates and samples capacity. CAD tools are then used to evaluate their capacity to assess die-to-die alpha and neutron SER variability.