Examples

All example source code described in this section lives under example/cusolverdx/ in the cuSolverDx package. Examples are grouped by topic in numbered directories:

• 00_Introduction/ — Introduction and batched Cholesky solve

• 01_Linear_Solve/ — Cholesky, LU, tridiagonal, and general linear solves (block, thread, and cluster execution)

• 02_Least_Squares/ — Batched least squares (block and thread)

• 03_Orthogonal_Factors/ — QR/LQ and Q application (block and thread)

• 04_Symmetric_Eigenvalues/ — HEEV, HTEV, HEGST, and HEGV (block and thread)

• 05_Singular_Value_Decomposition/ — BDSVD and GESVD (block and thread)

• 07_BLAS/ — Batched triangular solve (block and thread)

• 08_NVRTC/ — NVRTC runtime compilation

• 09_Performance/ — Performance benchmarks

• 10_Advanced/ — Blocked Cholesky (single-block and thread block cluster), batched TSQRHR using thread block clusters, regularized least squares

The cuSolverDx library provides block-level, thread-level, and cluster-level matrix decomposition examples covering basic to advanced operations with various functions, data types, layouts, and workflows.

• Introduction
• Linear Solve
• Least Squares
• Orthogonalization
• Symmetric Eigenvalue
• Singular Value Decomposition
• BLAS
• NVRTC
• Performance
• Advanced