Data transformation for the reduction of power and noise in CMOS structures
AssigneeTektronix, Inc.
InventorMichael S. Hagen
Life of this patent
Estimating the term…
Abstract
A data transformation algorithm is selectively applied to each data vector as it enters the pipelined structure. In a selection step, the algorithm compares the bit value of the new data vector with the corresponding bit values of the preceding data vector, and sums the number of logic transitions. The transformation algorithm is applied to the new data vector only if it would reduce the resulting number of transitions, otherwise the data vector is propagated unmodified. Bit inversion is a data transformation algorithm according to the present invention that provides up to a 50% reduction in the number of logic transitions.
Claims16
Circuitry for reducing power dissipation in a integrated circuit, comprising:
- a data latch for storing a first multibit digital signal;
- an input data latch for receiving a second multibit digital signal, said input data latch producing transformed and untransformed versions of said second multibit digital signal;
- a comparing circuit for comparing values of bits in respective bit positions of said first and second multibit digital signals and producing an output signal representative of a state transition at each bit position;
- summing circuitry coupled to said comparing circuitry for providing a sum of state transitions for each comparison of said multibit digital signals;
- a controller for determining if said sum exceeds a predetermined value, and producing a transformation vector signal in response thereto;
- a selection circuit coupled to said input data latch for selecting one of said transformed and said untransformed versions of said second multibit digital signal in response to said transformation vector signal.
A method for reducing power dissipation in a integrated circuit, comprising the steps of:
- storing a first multibit digital signal;
- receiving a second multibit digital signal;
- producing transformed and untransformed versions of said second multibit digital signal;
- comparing values of bits in respective bit positions of said first and second multibit digital signals and producing an output signal representative of a state transition at each bit position;
- providing a sum of state transitions for each comparison of said multibit digital signals;
- determining if said sum exceeds a predetermined value, and producing a transformation vector signal in response thereto;
- selecting one of said transformed and said untransformed versions of said second multibit digital signal in response to said transformation vector signal.
Dependencies are inferred from the claim text.