Prime Computer Inc.

Suren Irukulla

Bimal V Patel

Robert F Beckwith

Neil J Johnson

Nihar Mohapatra

Steven Schwartz

A pipelined central processor capable of executing both single-cycle instructions and multicycle instructions is provided. An instruction fetch stage of the processor includes an instruction cache memory and a prediction cache memory that are commonly addressed by a program counter register. The instruction cache memory stores instructions of a program being executed and microinstructions of a multicycle instruction interpreter. The prediction cache memory stores interpreter call predictions and interpreter entry addresses at the addresses of the multicycle intructions. When a call prediction occurs, the entry address of the instruction interpreter is loaded into the program counter register on the processing cycle immediately following the call prediction, and a return address is pushed onto a stack. The microinstructions of the interpreter are fetched sequentially from the instruction cache memory. When the interpreter is completed, the prediction cache memory makes a return prediction. The return address is transferred from the stack to the program counter register on the processing cycle immediately following the return prediction, and normal program flow is resumed. The prediction cache memory also stores branch instruction predictions and branch target addresses.

High-performance pipelined central processor for predicting the

A method and apparatus for radix-.beta. non-restoring division. The division process occurs in four phases. In a first phase, the input operands are transformed to produce a divisor lying in a designated numerical range. Next, a transitional phase involves generating an initial radix-.beta. quotient digit from the transformed numerator. An iterative phase of the process generates successive partial remainders according to a recursive method. From the sign and a single radix-.beta. digit of each of these partial remainders, the process generates a radix-.beta. quotient digit. Further, a fourth phase, which may run concurrently with the transitional and iterative phases, involves accumulating successively generated quotient digits to produce a final quotient value.

Method and apparatus for numerical division

A range transformation method for transforming the normalized divisor in a division calculation to a range wherein the transformed value differs from one by no more than the quantity 2.sup.-n. The method and apparatus generate the transform multiplier value from a first high order 'q' digits of the divisor and generate an out-of-range indicator signal from at least those same digits. The thus generated multiplier value is modified in response to the out-of-range indicator signal when an out-of-range condition is indicated. The apparatus employs a read-only-memory for enabling the generation of the transform multiplier value without requiring either large table look-up storage or multiplicative functions. As a result, various division methods requiring an initial transformation to provide a divisor which approaches one in value can be efficiently implemented.

Growing community of inventors

Holliston, MA, United States of America

Suren Irukulla