Abstract: A method includes forming a first transistor having a first gate dielectric thickness and a first source/drain extension depth, a second transistor having a second gate dielectric thickness and the first source/drain extension depth, and a third transistor having the second gate dielectric thickness and a second source/drain extension depth. The second source/drain extension depth is greater than the first source/drain extension depth. The second gate dielectric thickness is greater than the first gate dielectric thickness. The first transistor is used in a logic circuit. The third transistor is used in an I/O circuit. The second transistor is made without extra processing steps and is better than either the first or third transistor for coupling a power supply terminal to the logic circuit in a power-up mode and decoupling the power supply terminal from the logic circuit in a power-down mode.

Abstract: A method includes providing a wafer having a first die and a scribe grid, where the first die has die circuitry and a bond pad electrically connected to the die circuitry, and where the scribe grid has a scribe grid pad electrically connected to the die circuitry. The method further includes accessing the scribe grid pad to stimulate the die circuitry. A wafer includes a first die. The first die includes die circuitry, a plurality of conductive layers, and a bond pad electrically connected to the die circuitry via at least one conductive layer of the plurality of conductive layers. The wafer includes a scribe grid having a scribe grid pad, and an interconnect electrically connecting the scribe grid pad to the die circuitry. The plurality of die of the wafer can then be singulated, and at least one of the singulated die can be packaged.

Abstract: A circuit comprises a first plurality of transistors of a first channel length disposed along a speedpath, the first plurality of transistors providing a first timing performance. The circuit also comprises a second plurality of transistors of a second channel length having an expected equivalent functionality as the first plurality of transistors and disposed in parallel with the first plurality of transistors along the speedpath, wherein the second channel length is different from the first channel length. In addition, the circuit comprises an element configured to selectively replace the first plurality of transistors with the second plurality of transistors in response to a determination that the first timing performance of the first plurality of transistors fails a timing requirement of the speedpath. In one embodiment, the second channel length is a sub-minimal geometry with respect to the first channel length.

Abstract: A method includes forming a first transistor having a first gate dielectric thickness and a first source/drain extension depth, a second transistor having a second gate dielectric thickness and the first source/drain extension depth, and a third transistor having the second gate dielectric thickness and a second source/drain extension depth. The second source/drain extension depth is greater than the first source/drain extension depth. The second gate dielectric thickness is greater than the first gate dielectric thickness. The first transistor is used in a logic circuit. The third transistor is used in an I/O circuit. The second transistor is made without extra processing steps and is better than either the first or third transistor for coupling a power supply terminal to the logic circuit in a power-up mode and decoupling the power supply terminal from the logic circuit in a power-down mode.

Abstract: A method determines a body bias for a memory cell. A supply voltage is applied to the memory cell and a bit line is precharged to a voltage lower than the supply voltage. A programmable bias voltage circuit provides a bias voltage to the memory cell in response to values on its input. Initial test values for the input are used. The memory cell is tested to determine a pass or a fail condition of the memory cell. The initial values are retained as the input values if the memory cell passes. If the memory cell fails, the memory cell is tested at changed values for the input. If the changed input values result in the memory cell being in a pass condition, the programmable bias voltage circuit is configured, in non-volatile fashion, to have the changed input values.

Abstract: A circuit comprises a first plurality of transistors of a first channel length disposed along a speedpath, the first plurality of transistors providing a first timing performance. The circuit also comprises a second plurality of transistors of a second channel length having an expected equivalent functionality as the first plurality of transistors and disposed in parallel with the first plurality of transistors along the speedpath, wherein the second channel length is different from the first channel length. In addition, the circuit comprises an element configured to selectively replace the first plurality of transistors with the second plurality of transistors in response to a determination that the first timing performance of the first plurality of transistors fails a timing requirement of the speedpath. In one embodiment, the second channel length is a sub-minimal geometry with respect to the first channel length.

Abstract: A method determines a body bias for a memory cell. A supply voltage is applied to the memory cell and a bit line is precharged to a voltage lower than the supply voltage. A programmable bias voltage circuit provides a bias voltage to the memory cell in response to values on its input. Initial test values for the input are used. The memory cell is tested to determine a pass or a fail condition of the memory cell. The initial values are retained as the input values if the memory cell passes. If the memory cell fails, the memory cell is tested at changed values for the input. If the changed input values result in the memory cell being in a pass condition, the programmable bias voltage circuit is configured, in non-volatile fashion, to have the changed input values.

Abstract: A method includes providing a wafer having a first die and a scribe grid, where the first die has die circuitry and a bond pad electrically connected to the die circuitry, and where the scribe grid has a scribe grid pad electrically connected to the die circuitry. The method further includes accessing the scribe grid pad to stimulate the die circuitry. A wafer includes a first die. The first die includes die circuitry, a plurality of conductive layers, and a bond pad electrically connected to the die circuitry via at least one conductive layer of the plurality of conductive layers. The wafer includes a scribe grid having a scribe grid pad, and an interconnect electrically connecting the scribe grid pad to the die circuitry. The plurality of die of the wafer can then be singulated, and at least one of the singulated die can be packaged.

Abstract: A well-bias system dynamically adjusts well-bias set points to optimal levels across an integrated circuit (IC) for enhanced power savings and component reliability during a standby or low-power mode of operation. A controller within the IC determines if the chip power supply voltage will be reduced during an imminent standby or low power mode and sets a register controlling a negative well-bias set point for asserting well-bias to charge wells of the IC accordingly. To minimize leakage current without compromising reliability, the well-bias set point is set to (1) an optimal well-bias set point if a reduced supply voltage is to be applied to the IC, or (2) a minimum well-bias set point when a nominal or high supply voltage is to be applied to the IC.

Abstract: A flip-flop (10) comprises a first latch circuit (18), a second latch circuit (24), and a third latch circuit (26). The first latch circuit (18) is coupled to receive a clock signal and a first power supply voltage. The second latch circuit (24) is coupled to the first latch circuit (18) and receives the clock signal and the first power supply voltage. Preparatory to entering a low power mode, the third latch circuit (26) receives a second power supply voltage and is coupled to the second latch circuit (24) in response to a power down signal. During the low power mode, the first power supply voltage is removed from the first and second latch circuits (18, 24). When returning to a normal operating mode, the first power supply voltage is provided to the first and second latch circuits (18, 24), and the third latch circuit (26) is coupled to the first latch circuit (18) in response to a power restore signal.