International Business Machines Corporation

Steven Michael Burchfiel

Michael P Taborn

Roger Ned Bailey

David Terrence Matheny

James Henry Hesson

Geoffrey Francis Burns

A floating point arithmetic unit for performing independent multiply and add operations in the execution of a multiply-add instruction AC+B on three operands A, B, and C of p-bit precision includes a multiplier unit, a sticky collect unit, an adder unit, and a rounding unit. In addition, a risk condition detection unit provides detection of a risk condition corresponding to an occurrence of an imprecise resultant quantity prior to being rounded by the rounding unit. Upon detection of a risk condition, a trap is triggered and an extended sequence implementation unit carries out an extended multiply-add sequence and provides a multiply-add output having infinite precision prior to a final rounding. A floating point arithmetic method for performing independent multiply and add operations in the execution of a multiply-add instruction AC+B on three operands A, B, and C of p-bit precision is disclosed also.

Method and apparatus for performing high-precision multiply-add

A floating point processing system and method of operation are disclosed. Single word precision denormalized operands and also misaligned operands are detected while such operands are being loaded into the first stage of a pipelined floating point unit. Such operands are aligned to a double word boundary and/or single word operands are normalized by a processing stages inserted into the pipe ahead of the first stage of the pipelined floating point unit thereby entering a one stage delay. In this way, misaligned operands and single word denormalized operands can be processed without the need for example to cancel the instruction, execute a normalize or alignment instruction and then re-launch the original instruction.

Variable stage load path and method of operation

A system and method for denormalizing a floating point result is disclosed. Denormalized operands are capable of representing much smaller values than can be represented by a number normalized under the ANSI/IEEE standard 754-1985 that governs the representation of numbers in floating point notation to ensure uniformity among floating point notation users. The majority of results will be normalized operands and therefore the floating point unit pipeline is optimized to produce normalized results but contains wider exponent fields in order to represent values received as denormalized numbers. In order to return the result as a denormalized number with the smaller ANSI/IEEE exponent field, denormalization is accomplished by using the same pipeline resources by means of the floating point unit feedback path and uses one of the exponent equalizing alignment shifters and an incrementor in order to round the denormalized result. In this way, denormalized results can be provided without stopping the dispatching of instructions, without providing status bits in the register files and rename registers and without the hold signals often present in other floating point units to accomplish denormalization.

Denormalization system and method of operation

An apparatus for detecting sticky and a leading one includes first circuitry capable of detecting both sticky and a leading one. The apparatus further includes second circuitry that determines whether a sticky or a leading one detect is required. Depending upon that determination, the second circuitry controls the first circuitry to perform a sticky detection or a leading one detection. A method practiced by the apparatus includes the steps of detecting either sticky or a leading one utilizing the same circuitry, determining whether a sticky or a leading one detect is required, and controlling the circuitry in accordance with the determination of whether a sticky or a leading one detect is required.

Growing community of inventors

Austin, TX, United States of America

Steven Michael Burchfiel