A new class of low-rank short recurrences is proposed for nonsymmetric linear matrix equations, combining subspace projection with truncation and randomization to limit memory while accelerating convergence.
Isogeometric preconditioners based on fast solvers for the
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.NA 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A class of low-rank short recurrences for nonsymmetric linear matrix equations
A new class of low-rank short recurrences is proposed for nonsymmetric linear matrix equations, combining subspace projection with truncation and randomization to limit memory while accelerating convergence.