Abstract: A characterization array circuit provides accurate threshold voltage distribution values for process verification and improvement. The characterization array includes a circuit for imposing a fixed drain-source voltage and a constant channel current at individual devices within the array. A circuit for sensing the source voltage of the individual device is also included within the array. The statistical distribution of the threshold voltage is determined directly from the source voltage distribution by offsetting each source voltage by a value determined by completely characterizing one or more devices within the array. The resulting methodology avoids the necessity of otherwise characterizing each device within the array, thus reducing measurement time dramatically.

Abstract: A method 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.

Abstract: A IC wafer is fabricated using a process of interest to have a plurality of FET devices with different channel lengths (Leff) form a plurality of channel length groups. The threshold voltage (VT) is measured of a statistical sample of the FET devices in each channel length group at two different drain-to-source voltage (VDS). The mean of VT is calculated for each channel length and each VDS. A slope coefficient ? relating VT to Leff is calculated at each VDS. The total variance of VT is calculated at each VDS. Two equations at each VDS, each relating the total variance of VT to the variance of VT with respect to dopant levels and the square of the slope coefficient ? times the variance of Leff, are solved simultaneously to obtain the variance of VT with respect to dopant levels and the variance of Leff.

Abstract: A system and method for analyzing a memory element includes modeling the memory element using a simulation method and determining component response characteristics for components of the memory element. Safety regions are computed in a state space of the memory element, which indicate stable states. A transient analysis is performed to determine a path and time needed to reach one of the safety regions. Based on the path and time needed to reach one of the safety regions, a clock waveform or waveforms are determined which place a corresponding state in that safety region.

Abstract: A characterization array and method for determining threshold voltage variation rapidly provides accurate threshold voltage distribution values for process verification and improvement. The characterization array includes a circuit for imposing a fixed drain-source voltage and a constant channel current at individual devices within the array. A circuit for sensing the source voltage of the individual device is also included within the array. The statistical distribution of the threshold voltage is determined directly from the source voltage distribution by offsetting each source voltage by a value determined by completely characterizing one or more devices within the array. The resulting methodology avoids the necessity of otherwise characterizing each device within the array, thus reducing measurement time dramatically.

Abstract: A test structure for statistical characterization of local device mismatches contains densely populated SRAM devices arranged in a row/column addressable array that enables resource sharing of many devices. The test structure includes a built-in sensing mechanism to calibrate or null out sources of error, and current steering to avoid negative effects of current leakage along spurious paths. The gate and drain lines of each column are driven from both the top and bottom to minimizes parasitic effects. The system can handle a large number of devices while still providing high spatial resolution of current measurements.

Abstract: A circuit for computing moment pre-products for statistical analysis reduces data transfer volume for on-chip statistical measurements. The circuit calculates the sums of multiple exponentiations of outputs of one or more measurement circuits, thereby reducing the amount of data that must be transferred from a wafer. An integer scaling of the input data is arranged between zero and unity so that the exponentiations all similarly lie between zero and unity. The circuit can use look-up tables and adder/accumulators to accumulate the contributions of each measurement to each exponentiation, or use a multiplier arrangement to determine the contributions. The multipliers can be implemented in the adder/accumulators by clocking the adder/accumulators by corresponding counts determined from the measurement data and lower-order exponentiations. Ranges of the measurement values are determined by capturing maximum and minimum values using comparators as the measurements are input.

Abstract: An integrated circuit design has circuit macros made up of device cells. The cells are characterized by determining the leakage current dependency on various process, environmental and voltage parameters. When circuit macros are designed their leakage power is calculated using this data and multi-dimensional models for power and temperature distribution. Circuit macros are identified as timing-critical and timing-noncritical macros. Statistical methods are used to determine the average leakage sensitivities for the specific circuit macros designed. The designer uses the sensitivity data to determine how to redesign selected circuit macros to reduce leakage power. Reducing leakage power in these selected circuits may be used to reduce overall IC power or the improved power margins may be used in timing-critical circuits to increase performance while keeping power dissipation unchanged.

Abstract: A method for determining full chip leakage power first estimates leakage power and dynamic power for each circuit macro. The power supply voltage to each macro is first assumed to be nominal. The power dissipation for each macro is modeled as a current source whose value is the estimated power divided by the nominal power supply voltage. The power distribution network is modeled as a resistive grids. The thermal environment of the IC and its electronic package are modeled as multi dimensional grids of thermal elements. Algebraic multi-grid (AMG) methods are used to calculate updated circuit macro voltages and temperatures. The macro voltages and temperatures are updated and updated leakage and dynamic power dissipation are calculated. Iterations are continued until leakage power converges to a final value.