Abstract: Apparatuses, systems, and techniques to selectively use one or more neural network layers. In at least one embodiment, one or more neural network layers are selectively used based on, for example, one or more iteratively increasing neural network performance metrics.

Abstract: A computer-implemented method for executing a floating-point calculation where an exact value of an associated result cannot be expressed as a floating-point value is disclosed. The method involves: generating an estimate of the associated result and storing the estimate in memory; calculating an amount of error for the estimate; determining whether the amount of error is less than or equal to a threshold of error for the associated result; and if the amount of error is less than or equal to the threshold of error, then concluding that the estimate of the associated result is a correctly rounded result of the floating-point calculation; or if the amount of error is greater than the threshold of error, then testing whether the floating-point calculation constitutes an exception case.

Abstract: A computer-implemented method for executing a floating-point calculation where an exact value of an associated result cannot be expressed as a floating-point value is disclosed. The method involves: generating an estimate of the associated result and storing the estimate in memory; calculating an amount of error for the estimate; determining whether the amount of error is less than or equal to a threshold of error for the associated result; and if the amount of error is less than or equal to the threshold of error, then concluding that the estimate of the associated result is a correctly rounded result of the floating-point calculation; or if the amount of error is greater than the threshold of error, then testing whether the floating-point calculation constitutes an exception case.

Abstract: Transforming an integer comprises receiving the integer, where the integer can be expressed as a modular factorization. The modular factorization comprises one or more factors, where each factor has an exponent. The integer is expressed as a product of residues. A discrete logarithm of the integer is established from a sum corresponding to the product of residues. A value for an exponent of a factor is determined from the discrete logarithm. The integer is represented as the modular factorization comprising the one or more factors, where each factor has a value for the exponent.

Abstract: A system and method for efficient floating-point rounding in computer systems. A computer system may include at least one floating-point unit for floating-point arithmetic operations such as addition, subtraction, multiplication, division and square root. For the division operation, the constraints for the remainder may be relaxed in order to reduce the area for look-up tables. An extra internal precision bit may not be used. Only one quotient may be calculated, rather than two, further reducing needed hardware to perform the rounding. Comparison logic may be required that may add a couple of cycles to the rounding computation beyond the calculation of the remainder. However, the extra latency is much smaller than a second floating-point multiply accumulate latency.

Abstract: Determining a table output of a table representing a hierarchical tree for an integer valued function includes determining an address from a table input. A subset of a memory is selected according to the address, where the memory represents the hierarchical tree and the subset represents a subtree of the hierarchical tree. Bit fields are selected from the subset, and bits are extracted from the bit fields. A table output is determined from the extracted bits.

Abstract: A system and method for efficient floating-point rounding in computer systems. A computer system may include at least one floating-point unit for floating-point arithmetic operations such as addition, subtraction, multiplication, division and square root. For the division operation, the constraints for the remainder may be relaxed in order to reduce the area for look-up tables. An extra internal precision bit may not be used. Only one quotient may be calculated, rather than two, further reducing needed hardware to perform the rounding. Comparison logic may be required that may add a couple of cycles to the rounding computation beyond the calculation of the remainder. However, the extra latency is much smaller than a second FMAC latency.

Abstract: Determining a table output of a table representing a hierarchical tree for an integer valued function includes determining an address from a table input. A subset of a memory is selected according to the address, where the memory represents the hierarchical tree and the subset represents a subtree of the hierarchical tree. Bit fields are selected from the subset, and bits are extracted from the bit fields. A table output is determined from the extracted bits.

Abstract: Transforming an integer comprises receiving the integer, where the integer can be expressed as a modular factorization. The modular factorization comprises one or more factors, where each factor has an exponent. The integer is expressed as a product of residues. A discrete logarithm of the integer is established from a sum corresponding to the product of residues. A value for an exponent of a factor is determined from the discrete logarithm. The integer is represented as the modular factorization comprising the one or more factors, where each factor has a value for the exponent.