CS代写 DU-06709-001_v11.6 | March 2022

cuSOLVER Library
DU-06709-001_v11.6 | March 2022

Table of Contents

Copyright By PowCoder代写 加微信 powcoder

Chapter 1. Introduction………………………………………………………………………………………….. 1
1.1. cuSolverDN: Dense LAPACK……………………………………………………………………………………….2 1.2. cuSolverSP: Sparse LAPACK……………………………………………………………………………………… 2 1.3. cuSolverRF: Refactorization……………………………………………………………………………………….. 2 1.4. Naming Conventions…………………………………………………………………………………………………..3 1.5. Asynchronous Execution……………………………………………………………………………………………..4 1.6. Library Property………………………………………………………………………………………………………… 5 1.7. High Precision Package………………………………………………………………………………………………5
Chapter 2. Using the CUSOLVER API………………………………………………………………………. 6
2.1. General Description…………………………………………………………………………………………………… 6 2.1.2. Scalar Parameters………………………………………………………………………………………………. 6 2.1.3. Parallelism with Streams…………………………………………………………………………………….. 6 2.1.4. How to Link cusolver Library……………………………………………………………………………….. 6 2.1.5. Link Third-party LAPACK Library…………………………………………………………………………..7 2.1.6. Convention of info…………………………………………………………………………………………………7 2.1.7. Usage of _bufferSize…………………………………………………………………………………………….7
2.2. cuSolver Types Reference…………………………………………………………………………………………..8 2.2.1. cuSolverDN Types……………………………………………………………………………………………….. 8 2.2.1.1. cusolverDnHandle_t………………………………………………………………………………………..8 2.2.1.2. cublasFillMode_t…………………………………………………………………………………………….8 2.2.1.3. cublasOperation_t…………………………………………………………………………………………..8 2.2.1.4. cusolverEigType_t………………………………………………………………………………………….. 8 2.2.1.5. cusolverEigMode_t………………………………………………………………………………………….9 2.2.1.6. cusolverIRSRefinement_t……………………………………………………………………………….. 9 2.2.1.7. cusolverDnIRSParams_t………………………………………………………………………………..10 2.2.1.8. cusolverDnIRSInfos_t…………………………………………………………………………………….10 2.2.1.9. cusolverDnFunction_t……………………………………………………………………………………10 2.2.1.10. cusolverAlgMode_t………………………………………………………………………………………10 2.2.1.11. cusolverStatus_t………………………………………………………………………………………….11 2.2.2. cuSolverSP Types………………………………………………………………………………………………. 11 2.2.2.1. cusolverSpHandle_t……………………………………………………………………………………… 11 2.2.2.2. cusparseMatDescr_t……………………………………………………………………………………..11 2.2.2.3. cusolverStatus_t……………………………………………………………………………………………11 2.2.3. cuSolverRF Types………………………………………………………………………………………………. 12 2.2.3.1. cusolverRfHandle_t……………………………………………………………………………………….12
cuSOLVER Library DU-06709-001_v11.6 | ii

2.2.3.2. cusolverRfMatrixFormat_t……………………………………………………………………………..13 2.2.3.3. cusolverRfNumericBoostReport_t…………………………………………………………………. 13 2.2.3.4. cusolverRfResetValuesFastMode_t…………………………………………………………………13 2.2.3.5. cusolverRfFactorization_t………………………………………………………………………………13 2.2.3.6. cusolverRfTriangularSolve_t…………………………………………………………………………. 13 2.2.3.7. cusolverRfUnitDiagonal_t………………………………………………………………………………14 2.2.3.8. cusolverStatus_t……………………………………………………………………………………………14
2.3. cuSolver Formats Reference……………………………………………………………………………………..14 2.3.1. Index Base Format……………………………………………………………………………………………..14 2.3.2. Vector (Dense) Format………………………………………………………………………………………..14 2.3.3. Matrix (Dense) Format………………………………………………………………………………………..15 2.3.4. Matrix (CSR) Format………………………………………………………………………………………….. 15 2.3.5. Matrix (CSC) Format………………………………………………………………………………………….. 16
2.4. cuSolverDN: dense LAPACK Function Reference………………………………………………………..17 2.4.1. cuSolverDN Helper Function Reference……………………………………………………………….17 2.4.1.1. cusolverDnCreate()………………………………………………………………………………………. 17 2.4.1.2. cusolverDnDestroy()………………………………………………………………………………………18 2.4.1.3. cusolverDnSetStream()………………………………………………………………………………….18 2.4.1.4. cusolverDnGetStream()………………………………………………………………………………….18 2.4.1.5. cusolverDnCreateSyevjInfo()…………………………………………………………………………..19 2.4.1.6. cusolverDnDestroySyevjInfo()………………………………………………………………………… 19 2.4.1.7. cusolverDnXsyevjSetTolerance()……………………………………………………………………..19 2.4.1.8. cusolverDnXsyevjSetMaxSweeps()…………………………………………………………………..20 2.4.1.9. cusolverDnXsyevjSetSortEig()…………………………………………………………………………20 2.4.1.10. cusolverDnXsyevjGetResidual()……………………………………………………………………. 20 2.4.1.11. cusolverDnXsyevjGetSweeps()……………………………………………………………………… 21 2.4.1.12. cusolverDnCreateGesvdjInfo()……………………………………………………………………… 21 2.4.1.13. cusolverDnDestroyGesvdjInfo()……………………………………………………………………..22 2.4.1.14. cusolverDnXgesvdjSetTolerance()………………………………………………………………… 22 2.4.1.15. cusolverDnXgesvdjSetMaxSweeps()……………………………………………………………… 22 2.4.1.16. cusolverDnXgesvdjSetSortEig()……………………………………………………………………..23 2.4.1.17. cusolverDnXgesvdjGetResidual()…………………………………………………………………..23 2.4.1.18. cusolverDnXgesvdjGetSweeps()…………………………………………………………………….23 2.4.1.19. cusolverDnIRSParamsCreate()……………………………………………………………………..24 2.4.1.20. cusolverDnIRSParamsDestroy()……………………………………………………………………24 2.4.1.21. cusolverDnIRSParamsSetSolverPrecisions()………………………………………………….25 2.4.1.22. cusolverDnIRSParamsSetSolverMainPrecision()…………………………………………… 26 2.4.1.23. cusolverDnIRSParamsSetSolverLowestPrecision()……………………………………….. 27
cuSOLVER Library DU-06709-001_v11.6 | iii

2.4.1.24. cusolverDnIRSParamsSetRefinementSolver()………………………………………………..27 2.4.1.25. cusolverDnIRSParamsSetTol()…………………………………………………………………….. 29 2.4.1.26. cusolverDnIRSParamsSetTolInner()………………………………………………………………29 2.4.1.27. cusolverDnIRSParamsSetMaxIters()……………………………………………………………..30 2.4.1.28. cusolverDnIRSParamsSetMaxItersInner()…………………………………………………….. 31 2.4.1.29. cusolverDnIRSParamsEnableFallback()……………………………………………………….. 31 2.4.1.30. cusolverDnIRSParamsDisableFallback()………………………………………………………. 32 2.4.1.31. cusolverDnIRSParamsGetMaxIters()……………………………………………………………..32 2.4.1.32. cusolverDnIRSInfosCreate()………………………………………………………………………….33 2.4.1.33. cusolverDnIRSInfosDestroy()………………………………………………………………………..34 2.4.1.34. cusolverDnIRSInfosGetMaxIters()………………………………………………………………….34 2.4.1.35. cusolverDnIRSInfosGetNiters()……………………………………………………………………..35 2.4.1.36. cusolverDnIRSInfosGetOuterNiters()……………………………………………………………. 35 2.4.1.37. cusolverDnIRSInfosRequestResidual()…………………………………………………………..36 2.4.1.38. cusolverDnIRSInfosGetResidualHistory()……………………………………………………….36 2.4.1.39. cusolverDnCreateParams()…………………………………………………………………………. 37 2.4.1.40. cusolverDnDestroyParams()…………………………………………………………………………38 2.4.1.41. cusolverDnSetAdvOptions()…………………………………………………………………………..38
2.4.2. Dense Linear Solver Reference (legacy)……………………………………………………………….38 2.4.2.1. cusolverDnpotrf()…………………………………………………………………………………….. 38 2.4.2.2. cusolverDnPotrf()[DEPRECATED]……………………………………………………………………41 2.4.2.3. cusolverDnpotrs()……………………………………………………………………………………..43 2.4.2.4. cusolverDnPotrs()[DEPRECATED]…………………………………………………………………..45 2.4.2.5. cusolverDnpotri()………………………………………………………………………………………47 2.4.2.6. cusolverDngetrf()…………………………………………………………………………………….. 50 2.4.2.7. cusolverDnGetrf()[DEPRECATED]……………………………………………………………………52 2.4.2.8. cusolverDngetrs()……………………………………………………………………………………..55 2.4.2.9. cusolverDnGetrs()[DEPRECATED]…………………………………………………………………..57 2.4.2.10. cusolverDngesv()…………………………………………………………………………….58 2.4.2.11. cusolverDnIRSXgesv()…………………………………………………………………………………. 72 2.4.2.12. cusolverDngeqrf()……………………………………………………………………………………78 2.4.2.13. cusolverDnGeqrf()[DEPRECATED]…………………………………………………………………80 2.4.2.14. cusolverDngels()……………………………………………………………………………..82 2.4.2.15. cusolverDnIRSXgels()…………………………………………………………………………………..96 2.4.2.16. cusolverDnormqr()……………………………………………………………………………….. 101 2.4.2.17. cusolverDnorgqr()………………………………………………………………………………… 105 2.4.2.18. cusolverDnsytrf()………………………………………………………………………………….. 108 2.4.2.19. cusolverDnpotrfBatched()………………………………………………………………………111
cuSOLVER Library DU-06709-001_v11.6 | iv

2.4.2.20. cusolverDnpotrsBatched()…………………………………………………………………….. 113 2.4.3. Dense Eigenvalue Solver Reference (legacy)……………………………………………………….115 2.4.3.1. cusolverDngebrd()…………………………………………………………………………………..115 2.4.3.2. cusolverDnorgbr()………………………………………………………………………………….. 118 2.4.3.3. cusolverDnsytrd()…………………………………………………………………………………… 122 2.4.3.4. cusolverDnormtr()…………………………………………………………………………………..125 2.4.3.5. cusolverDnorgtr()…………………………………………………………………………………… 129 2.4.3.6. cusolverDngesvd()…………………………………………………………………………………..132 2.4.3.7. cusolverDnGesvd()[DEPRECATED]………………………………………………………………..135 2.4.3.8. cusolverDngesvdj()………………………………………………………………………………….139 2.4.3.9. cusolverDngesvdjBatched()…………………………………………………………………….. 144 2.4.3.10. cusolverDngesvdaStridedBatched()……………………………………………………….. 149 2.4.3.11. cusolverDnsyevd()………………………………………………………………………………… 155 2.4.3.12. cusolverDnSyevd()[DEPRECATED]……………………………………………………………….158 2.4.3.13. cusolverDnsyevdx()………………………………………………………………………………..161 2.4.3.14. cusolverDnSyevdx()[DEPRECATED]……………………………………………………………..166 2.4.3.15. cusolverDnsygvd()………………………………………………………………………………… 170 2.4.3.16. cusolverDnsygvdx()………………………………………………………………………………..175 2.4.3.17. cusolverDnsyevj()…………………………………………………………………………………..181 2.4.3.18. cusolverDnsygvj()…………………………………………………………………………………. 185 2.4.3.19. cusolverDnsyevjBatched()………………………………………………………………………190 2.4.4. Dense Linear Solver Reference (64-bit API)……………………………………………………….. 195 2.4.4.1. cusolverDnXpotrf()……………………………………………………………………………………….195 2.4.4.2. cusolverDnXpotrs()………………………………………………………………………………………197 2.4.4.3. cusolverDnXgetrf()……………………………………………………………………………………….199 2.4.4.4. cusolverDnXgetrs()………………………………………………………………………………………202 2.4.4.5. cusolverDnXgeqrf()………………………………………………………………………………………204 2.4.4.6. cusolverDnXsytrs()……………………………………………………………………………………… 206 2.4.4.7. cusolverDnXtrtri()……………………………………………………………………………………….. 208 2.4.5. Dense Eigenvalue Solver Reference (64-bit API)………………………………………………….210 2.4.5.1. cusolverDnXgesvd()……………………………………………………………………………………..211 2.4.5.2. cusolverDnXgesvdp()……………………………………………………………………………………214 2.4.5.3. cusolverDnXgesvdr()…………………………………………………………………………………….217 2.4.5.4. cusolverDnXsyevd()…………………………………………………………………………………….. 221 2.4.5.5. cusolverDnXsyevdx()…………………………………………………………………………………….224 2.5. cuSolverSP: sparse LAPACK Function Reference…………………………………………………….. 228 2.5.1. Helper Function Reference………………………………………………………………………………..228 2.5.1.1. cusolverSpCreate()………………………………………………………………………………………228
cuSOLVER Library DU-06709-001_v11.6 | v

2.5.1.2. cusolverSpDestroy()……………………………………………………………………………………. 229 2.5.1.3. cusolverSpSetStream()……………………………………………………………………………….. 229 2.5.1.4. cusolverSpXcsrissym()………………………………………………………………………………… 230
2.5.2. High Level Function Reference…………………………………………………………………………. 231 2.5.2.1. cusolverSpcsrlsvlu()………………………………………………………………………………..231 2.5.2.2. cusolverSpcsrlsvqr()………………………………………………………………………………. 234 2.5.2.3. cusolverSpcsrlsvchol()…………………………………………………………………………….237 2.5.2.4. cusolverSpcsrlsqvqr()…………………………………………………………………………….. 239 2.5.2

程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com