A Simulative Study of Correlated Error Propagation in Various Finite-Precision Arithmetics
John D. Marasa, David W. Matula
Abstract
John D. Marasa, David W. Matula
Abstract
The accumulated roundoff error incurred in long arithmetic computations involving a randomized mixture of addition, subtraction, multiplication, and division operations applied to an initial randomly generated data base is studied via simulation. Truncated and rounded floating-point arithmetic and truncated and rounded logarithmic arithmetic are simultaneously utilized for each of the computation sequences and the resulting roundoff error accumulations for these four systems are compared. The nature of the correlated errors incurred under various arithmetic operator mixes are discussed.
OpenAlex reports 21 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The accumulated roundoff error incurred in long arithmetic computations involving a randomized mixture of addition, subtraction, multiplication, and division operations applied to an initial randomly generated data base is studied via simulation. Truncated and rounded floating-point arithmetic and truncated and rounded logarithmic arithmetic are simultaneously utilized for each of the computation sequences and the resulting roundoff error accumulations for these four systems are compared. The nature of the correlated errors incurred under various arithmetic operator mixes are discussed.
Key concepts: Arithmetic, Multiplication (music), Round-off error, Saturation arithmetic, Arbitrary-precision arithmetic, Subtraction, Floating point, Computation