International Business Machines Corporation

Centre National De La Recherche Scientifique

Commissariat &#xc0; L'&#xc9;nergie Atomique Et Aux &#xc9;nergies Alternatives

Christophe J Layer

Silvia Melitta Mueller

Maarten Jakob Boersma

Klaus Michael Kroener

Markus Kaltenbach

Kerstin Claudia Schelm

Steven R Carlough

Jens Leenstra

Virgile Javerliac

Petra Leber

Michael Klein

Michael Klaus Kroener

Son Dao Trong

An arithmetic operation in a data processing unit, preferably by iterative digit accumulations, is proposed. An approximate result of the arithmetic operation is computed iteratively. Concurrently at least two supplementary values of the approximate result of the arithmetic operation are computed, and the final result selected from one of the values of the approximate result and the at least two supplementary values of the arithmetic operation depending on the results of the last iteration step.

Arithmetic operation in a data processing system

A tool for supporting vector operations in a scalar data path. The tool determines a location for a split in the scalar mode configuration to enable vector mode operations. The tool determines the number of coarse shift multiplexers in conflict at a bit position receiving data from both the left half and the right half of the vector mode configuration. The tool duplicates a coarse shift multiplexer in conflict at the bit position receiving data from both the left half and the right half of the vector mode configuration. The tool duplicates an intermediate data signal in conflict at a signal position receiving data from both the left half and the right half of the vector mode configuration. The tool receives a control signal to split the scalar mode configuration and shift the leading zeros across the left half and the right half of the vector mode configuration.

Splitable and scalable normalizer for vector data

A memory array including: a first volatile memory cell including first and second cross-coupled inverters between first and second storage nodes; a first non-volatile memory cell including at least one resistive element that can be programmed to take one of at least two resistive states; and a control circuit adapted to couple the first non-volatile memory cell to the first and second storage nodes in order to generate a current for programming the resistive state of the at least one resistive element.

Method and circuit for programming non-volatile memory cells of a volatile/non-volatile memory array

A memory array including: a plurality of volatile memory cells, each including a latch; and a plurality of non-volatile memory cells, each including at least one resistive element that can be programmed by the direction of a current passed therethrough in order to take at least two resistive states, each of the non-volatile memory cells being associated with a corresponding cell from the volatile memory cells.

Memory provided with associated volatile and non-volatile memory cells

A method for operating a fused-multiply-add pipeline in a floating-point unit of a processor is disclosed. A multiplication is initially performed between a first operand and a second operand in a multiplier block to obtain a set of partial product results. The partial product results are sent to a carry-save adder block. A partial product reduction is performed on the partial product results to generate a carry-save result having a sum term and a carry term. The carry-save result is then formatted to generate a carry-out bit. The carry-save result is added to a third operand to generate a final result.

Fused multiply add pipeline

A hardware circuit component configured to support vector operations in a scalar data path. The hardware circuit component configured to operate in a vector mode configuration and in a scalar mode configuration. The hardware circuit component configured to split the scalar mode configuration into a left half and a right half of the vector mode configuration. The hardware circuit component configured to perform one or more bit shifts over one or more stages of interconnected multiplexers in the vector mode configuration. The hardware circuit component configured to include duplicated coarse shift multiplexers at bit positions that receive data from both the left half and the right half of the vector mode configuration, resulting in one or more coarse shift multiplexers sharing the bit position.

A fused multiply-adder is disclosed. The fused multiply-adder includes a Booth encoder, a fraction multiplier, a carry corrector, and an adder. The Booth encoder initially encodes a first operand. The fraction multiplier multiplies the Booth-encoded first operand by a second operand to produce partial products, and then reduces the partial products into a set of redundant sum and carry vectors. The carry corrector then generates a carry correction factor for correcting the carry vectors. The adder adds the redundant sum and carry vectors and the carry correction factor to a third operand to yield a final result.

Fused multiply-adder with booth-encoding

According to one aspect of the present disclosure, a method and technique for encoding densely packed decimals is disclosed. The method includes: executing a floating point instruction configured to perform a floating point operation on decimal data in a binary coded decimal (BCD) format; determining whether a result of the operation includes a rounded mantissa overflow; and responsive to determining that the result of the operation includes a rounded mantissa overflow, compressing a result of the operation from the BCD-formatted decimal data to decimal data in a densely packed decimal (DPD) format by shifting select bit values of the BCD formatted decimal data by one digit to select bit positions in the DPD format.

Encoding densely packed decimals

Techniques for reducing issue-to-issue latency by reversing processing order in half-pumped single instruction multiple data (SIMD) execution units are described. In one embodiment a processor functional unit is provided comprising a frontend unit, and execution core unit, a backend unit, an execution order control signal unit, a first interconnect coupled between and output and an input of the execution core unit and a second interconnect coupled between an output of the backend unit and an input of the frontend unit. In operation, the execution order control signal unit generates a forwarding order control signal based on the parity of an applied clock signal on reception of a first vector instruction. This control signal is in turn used to selectively forward first and second portions of an execution result of the first vector instruction via the interconnects for use in the execution of a dependent second vector instruction.

Reversing processing order in half-pumped SIMD execution units to achieve K cycle issue-to-issue latency

Mechanisms for utilizing a reduced lookup table circuit to perform an operation in a data processing device are provided. A first input value is input for selecting a subset of values from the reduced lookup table circuit. The reduced lookup table circuit stores only boundary cell values from a fully filled lookup table corresponding to the reduced lookup table circuit. The subset of values comprises only a subset of boundary cell values corresponding to the first input value. A second value is input and a comparison, by the reduced lookup table circuit, of the second value to each of the boundary cell values in the subset of boundary cell values is performed. The reduced lookup table circuit outputs an output value based on results of the comparison of the second value to each of the boundary cell values in the subset of boundary cell values.

Growing community of inventors

Grenoble, France

Christophe J Layer