We define “reproducibility” as getting bitwise identical results from multiple runs of the same
program, perhaps with different hardware resources or other changes that should not affect …

Reproducibility means getting the bitwise identical floating point results from multiple runs of
the same program, which plays an essential role in debugging and correctness checking in …

Abstract The Preconditioned Conjugate Gradient method is often used in numerical
simulations. While being widely used, the solver is also known for its lack of accuracy while …

The Preconditioned Conjugate Gradient method is often employed for the solution of linear
systems of equations arising in numerical simulations of physical phenomena. While being …

Reproducibility of a deep-learning fully convolutional neural network is evaluated by training
several times the same network on identical conditions (database, hyperparameters, and …

We propose a reproducible variant of the unblocked LU factorization for graphics processor
units (GPUs). For this purpose, we build upon Level-1/2 BLAS kernels that deliver correctly …

Проблема влияния ошибок округления возникает в большом количестве задач в
различных областях знаний, включая вычислительную математику, математическую …

Введение. Для решения задач моделирования, критичных к ошибкам округления
(включая задачи вычислительной математики, математической физики, оптимального …

Prédire la propagation du bruit aéroacoustique est une tâche difficile en présence
d'écoulements moyens complexes et d'effets géométriques d'installation. La conception des …

In this article, we address the problem of reproducibility of the blocked LU factorization on
GPUs due to cancellations and rounding errors when dealing with floating-point arithmetic …