Ասցատրյան Հ., Գիչունց Է., Асцатрян Г., Гичунц Э.
We consider some important issues related to the solution of linear system of equations that arise in multi-processor and graphics processing unit architecture. A more effective method for solving a linear system of equations is considered through the LU factorization. Investigations are conducted in case of general complex matrices, because for those matrices the random butterfly transformation is used. The paper presents performances of several ways of solving methods on the graphic processor NVIDIA K40c.
oai:arar.sci.am:258868
ՀՀ ԳԱԱ Հիմնարար գիտական գրադարան
Dec 8, 2023
Jul 24, 2020
27
https://arar.sci.am/publication/281995
Edition name | Date |
---|---|
Performances of Methods for Solving a Linear System of Equations in the Architecture of GPU Accelerator | Dec 8, 2023 |
Hrachya V. Astsatryan Edita E. Gichunts
Vladimir G. Sahakyan Yuri H. Shoukourian Hrachya V. Astsatryan
Hrachya V. Astsatryan Edita E. Gichunts
Seda N. Manukian
E. P. Serrano M. I. Troparevsky M. A. Fabio Գլխավոր խմբ․՝ Մ․ Մ․ Ջրբաշյան (1966-1994) Ռ․ Վ․ Համբարձումյան (1994-2009) Ա․ Ա․ Սահակյան (2010-)
Hrachya V. Astsatryan Edita E. Gichunts