US 6,148,314 AGranted

Round increment in an adder circuit

AssigneeArm Limited

InventorsDavid Terrence Matheny, David Vivian Jaggar, David James Seal

Life of this patent

Estimating the term…

Loading patent…

Abstract

A floating point unit is described that performs addition operations. An adder 16 within the floating point unit receives a first input and a second input to generate a sum. This sum is subject to subsequent normalization by a normalizer 60 and rounding by an incrementer 64. If an operation is performed that is immediately followed by an addition operation using the result of the preceding operation, then the normalized but unrounded sum is fed back to the adder 16 together with an indication of its rounding requirement. This rounding requirement can be performed by the adder 16 in parallel with the execution of the following addition by using the carry-in bit of the adder 16 to apply any increment required to rounding of the preceding result.

Claims15

Dependent claims

Apparatus for performing a floating point addition in which a first operand is added to a second operand, said apparatus comprising:

an alignment-shifter for floating-point-significance-aligning said first operand to generate an aligned first operand;
an adder having a first-adder-input and a second-adder-input for adding said aligned first operand input to said first-adder-input to said second operand input to said second-adder-input to generate a sum;
a normalizer for normalizing said sum to generate a normalized sum;
a rounding-increment-determination-circuit for generating a rounding signal indicating whether or not said normalized sum should be subject to a rounding-increment;
and
an incrementer responsive to said round signal for incrementing said normalized sum to generate a rounded sum;
wherein
when performing a succeeding floating point addition that uses as one input operand a result of said floating point addition, said normalized sum is fed back as said second operand to said second-adder-input and said rounding signal is used to generate a carry-in input to said adder such that said adder gives effect to any rounding-increment for said normalized sum.

5 more elements

Dependencies are inferred from the claim text.