Abstract: One embodiment of the invention comprises, in each clock zone of a central processing unit, at least one sensor that generates a power signal indicative of a power supply voltage within the clock zone, a clock generator for providing a variable frequency clock to the clock zone, and a controller for controlling an operating frequency of the clock generator in response to the power signal and in response to frequency adjustment communications from other clock zones.

Abstract: Systems and methods of edge calibration for synchronous data transfer between clock domains are disclosed. An exemplary method may comprise comparing a drive clock signal to a receive clock signal, generating a select clock signal, and configuring a data path based a least in part on the select clock signal for synchronous data transfer between clock domains so that data arrives in an early clock domain at the desired logical clock cycle.

Abstract: Systems and methods for implanting synchronous data transfer between clock domains are disclosed. An exemplary system may comprise an adaptable data path having an input for receiving a signal from a first clock domain and an output in a second clock domain. A controller is operatively associated with the adaptable data path. The controller is responsive to operating parameters to configure the adaptable data path to align a logical clock pulse on the signal received from the first clock domain with the same logical clock pulse in the second clock domain based on a measured delay between the first and second clock domains.

Abstract: Systems and methods for implementing count calibration for synchronous data transfer between clock domains are disclosed. An exemplary system may include a count calibration circuit for determining latency between an early clock domain and a late clock domain. The system may also include a data path configurable for synchronous data transfer between clock domains based at least in part on the latency.

Abstract: In one embodiment, a central processing unit (CPU) includes multiple clock zones. Each clock zone includes at least one sensor that generates a signal indicative of a power supply voltage within the clock zone, a clock generator for providing a variable frequency clock to the clock zone, a first controller for controlling a frequency of operation of the clock generator in response to the at least one sensor, wherein the first controller further controls the frequency of operation in response to communication of frequency adjustments from first controllers in other clock zones within one cycle of latency, and a second controller that provides an overdrive signal, that is combined with adjustment signals from the first controller for the clock generator, in response to communication of frequency adjustments from other clock zones beyond one cycle of latency.

Abstract: In a preferred embodiment, the invention provides a method and system for allowing a frequency synthesizer to function despite long delays. A first and second phase comparator, each with at least three inputs and an output are preset to a predetermined logical value by a first control circuit. A first signal is connected to an input of the first and second phase comparators. A second signal is connected to a second input of the second phase comparator and to the input of a programmable dead zone delay circuit. The output of the programmable dead zone delay circuit is connected to a second input of the first phase comparator. A preset value, determined by the first control circuit, is presented on the outputs of the first and second phase comparators. Until metastability is resolved, these outputs retain a valid fail-safe default.

Abstract: A method of compacting an instruction queue in an out of order processor includes determining the number of invalid instructions below and including each row in the queue, by counting invalid bits or validity indicators associated with rows below and up to the current row. For each row, multiplexor select signals are generated from the flat vector counts for the N rows above and including the present row, and from the validity indicators associated with the N rows, where N is a predetermined value. A multiplexor associated with a particular row selects one of the N rows according to the select value, and moves or passes the instruction held in the selected row to the present row. A row's select value is determined by forming a diagonal from the N count vectors corresponding to the N rows above and including the present row, and logically ANDing, each diagonal bit with the valid bit associated with the same row. Each row's count vector is determined in two stages.

Abstract: A method of compacting an instruction queue in an out of order processor includes determining the number of invalid instructions below and including each row in the queue, by counting invalid bits or validity indicators associated with rows below and up to the current row. For each row, multiplexor select signals are generated from the flat vector counts for the N rows above and including the present row, and from the validity indicators associated with the N rows, where N is a predetermined value. A multiplexor associated with a particular row selects one of the N rows according to the select value, and moves or passes the instruction held in the selected row to the present row. A row's select value is determined by forming a diagonal from the N count vectors corresponding to the N rows above and including the present row, and logically ANDing, each diagonal bit with the valid bit associated with the same row. Each row's count vector is determined in two stages.

Abstract: A computer hardware system is disclosed for determining during a single clock cycle whether a data buffer having a plurality of entries can accept additional data. The system has multiple stages, having one or more adders/encoders that process the data buffer entries' valid bits in parallel. Groups of entries are associated with first-stage adders/encoders. Valid bits and their complements for entries in each group are received into multiple first-stage adders that compute and output encoded values indicating the number of available entries within each group, or first-stage totals. The adders also encode the first-stage totals such that a saturated count corresponds to a pre-charged state of the first-stage adder. The first-stage totals are then sent to additional stages having adders/encoders that are substantially the same as the first-stage adders/encoders. The additional-stage adders combine the encoded totals from prior stages and determine whether the buffer has available entries.

Abstract: A computer hardware system is disclosed for determining during a single clock cycle whether a data buffer having a plurality of entries can accept additional data. The system has multiple stages, having one or more adders/encoders that process the data buffer entries' valid bits in parallel. Groups of entries are associated with first-stage adders/encoders. Valid bits and their complements for entries in each group are received into multiple first-stage adders that compute and output encoded values indicating the 8 number of available entries within each group, or first-stage totals. The adders also encode the first-stage totals such that a saturated count corresponds to a pre-charged state of the first-stage adder. The first-stage totals are then sent to additional stages having adders/encoders that are substantially the same as the first-stage adders/encoders. The additional-stage adders combine the encoded totals from prior stages and determine whether the buffer has available entries.

Abstract: A pipelined floating point processor including an add pipe for performing floating point additions is described. The add pipe includes a circuit to predict a normalization shift amount from examination of input operands, a circuit to determine the "Sticky bit" from the input operands, and a rounding adder which adds a pair of operands and rounds the result in a single pipeline stage operation. The rounding adder incorporates effects due to rounding in select logic for a series of carry select adders. The adder also aligns the datapath to permit economical storage and retrieval of floating point and integer operands for floating point or conversions operations. The floating point processor also includes in the adder pipeline a divider circuit include a quotient register having overflow quotient bit positions to detect the end of a division operation.

Abstract: A pipelined floating point processor including an add pipe for performing floating point additions is described. The add pipe includes a circuit to predict a normalization shift amount from examination of input operands, a circuit to determine the "Sticky bit" from the input operands, and a rounding adder which adds a pair of operands and rounds the result in a single pipeline stage operation. The rounding adder incorporates effects due to rounding in select logic for a series of carry select adders. The adder also aligns the datapath to permit economical storage and retrieval of floating point and integer operands for floating point or conversions operations. The floating point processor also includes in the adder pipeline a divider circuit include a quotient register having overflow quotient bit positions to detect the end of a division operation.

Abstract: A pipelined floating point processor including an add pipe for performing floating point additions is described. The add pipe includes a circuit to predict a normalization shift amount from examination of input operands, a circuit to determine the "Sticky bit" from the input operands, and a rounding adder which adds a pair of operands and rounds the result in a single pipeline stage operation. The rounding adder incorporates effects due to rounding in select logic for a series of carry select adders. The adder also aligns the datapath to permit economical storage and retrieval of floating point and integer operands for floating point or conversions operations. The floating point processor also includes in the adder pipeline a divider circuit include a quotient register having overflow quotient bit positions to detect the end of a division operation.

Abstract: A pipelined floating point processor including an add pipe for performing floating point additions is described. The add pipe includes a circuit to predict a normalization shift amount from examination of input operands, a circuit to determine the "Sticky bit" from the input operands, and a rounding adder which adds a pair of operands and rounds the result in a single pipeline stage operation. The rounding adder incorporates effects due to rounding in select logic for a series of carry select adders. The adder also aligns the datapath to permit economical storage and retrieval of floating point and integer operands for floating point or conversions operations. The floating point processor also includes in the adder pipeline a divider circuit include a quotient register having overflow quotient bit positions to detect the end of a division operation.

Abstract: A pipelined floating point processor including an add pipe for performing floating point additions described. The add pipe includes a circuit to predict a normalization shift amount from examination of input operands, a circuit to determine the "Sticky bit" from the input operands, and a rounding adder which adds a pair of operands and rounds the result in a single pipeline stage operation. The rounding adder incorporates effects due to rounding in select logic for a series of carry select adders. The adder also aligns the datapath to permit economical storage and retrieval of floating point and interger operands for floating point or conversions operations. The floating point processor also includes in the adder pipeline a divider circuit include a quotient register having overflow quotient bit positions to detect the end of a division operation.