Abstract: An apparatus to facilitate hardware enhancements for double precision systolic support is disclosed. The apparatus includes matrix acceleration hardware having double-precision (DP) matrix multiplication circuitry including a multiplier circuits to multiply pairs of input source operands in a DP floating-point format; adders to receive multiplier outputs from the multiplier circuits and accumulate the multiplier outputs in a high precision intermediate format; an accumulator circuit to accumulate adder outputs from the adders with at least one of a third global source operand on a first pass of the DP matrix multiplication circuitry or an intermediate result from the first pass on a second pass of the DP matrix multiplication circuitry, wherein the accumulator circuit to generate an accumulator output in the high precision intermediate format; and a down conversion and rounding circuit to down convert and round an output of the second pass as final result in the DP floating-point format.

Abstract: A bit line that has a feedback path from the bit line to a storage cell on the bit line is provided. The feedback path allows the bit line to discharge through a discharge device that is connected to a non-discharging local bit line. Further, a discharge device capable of discharging a global bit line even when a storage cell connected to the discharge device is not being evaluated is provided. Further, a method to perform a memory array operation by discharging a bit line using multiple discharge devices is provided.

Abstract: A bit line that has a feedback path from the bit line to a storage cell on the bit line is provided. The feedback path allows the bit line to discharge through a discharge device that is connected to a non-discharging local bit line. Further, a discharge device capable of discharging a global bit line even when a storage cell connected to the discharge device is not being evaluated is provided. Further, a method to perform a memory array operation by discharging a bit line using multiple discharge devices is provided.

Abstract: A multiplexer based adder circuit. The novel adder design is suitable for a number of bit sizes, but in one exemplary embodiment is a 64-bit adder. A complete 16-bit scaled adder is taught. The adder circuit is efficient and reconfigurable in that the adder can be partitioned to support a variety of data formats. The adder can add two 64-bit operands, four 32-bit operands, eight 16-bit operands, or sixteen 8-bit operands. The reconfigurability of the adder for different word sizes is achieved using only a small number of control signals for partitioning without increasing the adder size or reducing its speed. The novel adder circuit is designed using multiplexer circuits and two input inverted logic gates making the adder very fast. The adder design recognizes that pass transistor based multiplexer circuits and inverted logic gates are the fastest circuit elements for standard CMOS logic.

Abstract: A partitioned multiplier circuit which is designed for high speed operations. The multiplier of the present invention can perform one 32×32 bit multiplication, two 16×16 bit multiplications (simultaneously) or four 8×8 bit multiplications (simultaneously) depending on input partitioning signals. The time required to perform either the 32×32 bit or the 16×16 bit or the 8×8 bit multiplications is constant. Therefore, multiplication results are available with a constant latency regardless of operand bit-size. In one embodiment, the latency is two clock cycles but the multiplier circuit has a throughput of one clock cycle due to pipelining. The input operands can be signed or unsigned. The hardware is partitioned without any significant increase in the delay or area and the multiplier can provide six different modes of operation.

Abstract: An improved Booth encoder/selector circuit having an optimized critical path. The Booth encoder has a number of inverters coupled to several of the input multiplier bits. The inverted/non-inverted multiplier bits are then fed as inputs to NAND gates as well as a series of pass gates. The outputs of the pass gates are then fed as inputs to other NAND gates. The output from the NAND gates serve as control signals for controlling the Booth selector. The Booth selector is comprised of inverters and pass gates. Multiplicand bits are input to the pass gates. The control signals generated by the Booth encoder are selectively coupled to the inverters and pass gates such that they control which one of a plurality of multiplicand bits are selected for output. Basically, the Booth selector functions as a multiplexer whereby one of the following is output: the multiplicand bit is multiplied by zero, multiplied by one, multiplied by negative one, multiplied by two, or multiplied by negative two.

Abstract: A pipelined data path architecture for use, in one embodiment, in a multimedia processor. The data path architecture requires a maximum of two execution pipestages to perform all instructions including wide data format multiply instructions and specially adapted multimedia instructions, such as the sum of absolute differences (SABD) instruction and other multiply with add (MADD) instructions. The data path architecture includes two wide data format input registers that feed four partitioned 32×32 multiplier circuits. Within two pipestages, the multiply circuit can perform one 128×128 multiply operation, four 32×32 multiply operations, eight 16×16 multiply operations or sixteen 8×8 multiply operations in parallel. The multiply circuit contains a compressor tree which generates a 256-bit sum and a 256-bit carry vector. These vectors are supplied to four 64-bit carry propagate adder circuits which generate the multiply results.

Abstract: A partitioned shift right logic circuit that is programmable and contains rounding support. The circuit of the present invention accepts a 32-bit value and a shift amount and then performs a right shift operation on the 32-bits and automatically rounds the result(s). Signed or unsigned values can be accepted. The right shift circuit is partitioned so that the 32-bit value can represent: (1) a single 32-bit number; or (2) two 16-bit values. A 1 bit selection input indicates the particular partition format. In operation, if the input value is not negative, then one (“1”) is added at the guard bit position and a right shift with truncate is performed. If the input is negative and the guard bit is zero, then no addition is done and a right shift with truncate is performed. If the input is negative and the guard bit is one and the sticky bit is zero, then no addition is done and a right shift with truncate is performed.