Using Fuzzy-Jbit location of floating-point multiply-accumulate results

A method to be performed by a processor, the method comprising:

• decoding, using decode circuitry, an instruction specifying locations of first, second, and third floating-point (FP) operands and an opcode calling for accumulating a FP product of the first and second FP operands with previous contents of the third FP operand;
• using execution circuitry to, in a first cycle, generate the FP product having a Fuzzy-Jbit format comprising a sign bit, a 9-bit exponent, and a 25-bit mantissa having two bits to allow two possible positions for a Jbit;
• and
• using the execution circuitry, in a second cycle, to accumulate the FP product with the third FP operand, while concurrently, based on Jbit positions of the FP product and the third FP operand, to determine an exponent adjustment and a mantissa shift control of a result of the accumulation, wherein performing the exponent adjustment concurrently with the accumulation reduces an impact of the exponent adjustment on a critical timing path, and enhances an ability to perform the accumulation in one cycle.

A non-transitory machine-readable medium containing instructions to which a processor is to respond by:

• decoding, using decode circuitry, an instruction specifying locations of first, second, and third floating-point (FP) operands and an opcode calling for accumulating a FP product of the first and second FP operands with previous contents of the third FP operand;
• using execution circuitry to, in a first cycle, generate the FP product having a Fuzzy-Jbit format comprising a sign bit, a 9-bit exponent, and a 25-bit mantissa having two bits to allow two possible positions for a Jbit;
• and
• using the execution circuitry, in a second cycle, to accumulate the FP product with the third FP operand, while concurrently, based on Jbit positions of the FP product and the third FP operand, to determine an exponent adjustment and a mantissa shift control of a result of the accumulation, wherein performing the exponent adjustment concurrently with the accumulation reduces an impact of the exponent adjustment on a critical timing path, and enhances an ability to perform the accumulation in one cycle.