warp.tile_cholesky_inplace

warp.tile_cholesky_inplace(

A: Tile[Float, tuple[int, int]],

) → None

Kernel

Compute the Cholesky factorization L of a matrix A.

L is lower triangular and satisfies LL^T = A.

Only the lower triangular portion of A is used for the decomposition; the upper triangular part may be left unspecified.

Note: This inplace variant does not support automatic differentiation (adjoint computation), but offers improved performance and uses half the shared memory compared to the standard version.

Supported datatypes are:

• float32

• float64

param A:

A square, symmetric positive-definite, matrix. Only the lower triangular part of A is replaced by L, such that LL^T = A; the upper part is untouched.