1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
|
*************
Introduction
*************
.. toctree::
:maxdepth: 4
.. contents:: Table of contents
:local:
:backlinks: top
Library overview
==========================
hipSOLVER is an open-source marshalling library for `LAPACK routines <https://www.netlib.org/lapack/explore-html/modules.html>`_ on the GPU.
It sits between a backend library and the user application, marshalling inputs to and outputs from the backend library so that the user
application remains unchanged when using different backends. Currently, two backend libraries are supported by hipSOLVER: NVIDIA's `cuSOLVER
library <https://developer.nvidia.com/cusolver>`_ and AMD's open-source `rocSOLVER library <https://github.com/ROCmSoftwarePlatform/rocSOLVER>`_.
The :ref:`regular hipSOLVER API <library_api>` is a thin wrapper layer around the different backends. As such, it is not expected to introduce
significant overhead. However, its main purpose is portability, so when performance is critical, directly using the library backend corresponding
to the given architecture is recommended.
Another purpose of hipSOLVER is to facilitate the translation of cuSOLVER applications to
`AMD's open source ROCm platform <https://rocmdocs.amd.com/en/latest/index.html>`_ ecosystem. To that end, hipSOLVER also includes a
:ref:`compatibility API <library_compat>` with method signatures that match exactly those of cuSOLVER. For more details see the
section :ref:`usage_label`.
Currently implemented functionality
====================================
The hipSOLVER library remains in active development. New features are being continuously added, with new functionality documented at each
`release of the ROCm platform <https://rocmdocs.amd.com/en/latest/Current_Release_Notes/Current-Release-Notes.html>`_.
The following tables summarize the wrapper functions that are implemented in the regular API for the different supported precisions in
hipSOLVER's latest release. Most of these functions have a corresponding version in the compatibility API, where applicable.
LAPACK auxiliary functions
----------------------------
.. csv-table:: Orthonormal matrices
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXorgbr_bufferSize <orgbr_bufferSize>`, x, x, ,
:ref:`hipsolverXorgbr <orgbr>`, x, x, ,
:ref:`hipsolverXorgqr_bufferSize <orgqr_bufferSize>`, x, x, ,
:ref:`hipsolverXorgqr <orgqr>`, x, x, ,
:ref:`hipsolverXorgtr_bufferSize <orgtr_bufferSize>`, x, x, ,
:ref:`hipsolverXorgtr <orgtr>`, x, x, ,
:ref:`hipsolverXormqr_bufferSize <ormqr_bufferSize>`, x, x, ,
:ref:`hipsolverXormqr <ormqr>`, x, x, ,
:ref:`hipsolverXormtr_bufferSize <ormtr_bufferSize>`, x, x, ,
:ref:`hipsolverXormtr <ormtr>`, x, x, ,
.. csv-table:: Unitary matrices
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXungbr_bufferSize <ungbr_bufferSize>`, , , x, x
:ref:`hipsolverXungbr <ungbr>`, , , x, x
:ref:`hipsolverXungqr_bufferSize <ungqr_bufferSize>`, , , x, x
:ref:`hipsolverXungqr <ungqr>`, , , x, x
:ref:`hipsolverXungtr_bufferSize <ungtr_bufferSize>`, , , x, x
:ref:`hipsolverXungtr <ungtr>`, , , x, x
:ref:`hipsolverXunmqr_bufferSize <unmqr_bufferSize>`, , , x, x
:ref:`hipsolverXunmqr <unmqr>`, , , x, x
:ref:`hipsolverXunmtr_bufferSize <unmtr_bufferSize>`, , , x, x
:ref:`hipsolverXunmtr <unmtr>`, , , x, x
LAPACK main functions
----------------------------
.. csv-table:: Triangular factorizations
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXpotrf_bufferSize <potrf_bufferSize>`, x, x, x, x
:ref:`hipsolverXpotrf <potrf>`, x, x, x, x
:ref:`hipsolverXpotrfBatched_bufferSize <potrf_batched_bufferSize>`, x, x, x, x
:ref:`hipsolverXpotrfBatched <potrf_batched>`, x, x, x, x
:ref:`hipsolverXgetrf_bufferSize <getrf_bufferSize>`, x, x, x, x
:ref:`hipsolverXgetrf <getrf>`, x, x, x, x
:ref:`hipsolverXsytrf_bufferSize <sytrf_bufferSize>`, x, x, x, x
:ref:`hipsolverXsytrf <sytrf>`, x, x, x, x
.. csv-table:: Orthogonal factorizations
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXgeqrf_bufferSize <geqrf_bufferSize>`, x, x, x, x
:ref:`hipsolverXgeqrf <geqrf>`, x, x, x, x
.. csv-table:: Problem and matrix reductions
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXsytrd_bufferSize <sytrd_bufferSize>`, x, x, ,
:ref:`hipsolverXsytrd <sytrd>`, x, x, ,
:ref:`hipsolverXhetrd_bufferSize <hetrd_bufferSize>`, , , x, x
:ref:`hipsolverXhetrd <hetrd>`, , , x, x
:ref:`hipsolverXgebrd_bufferSize <gebrd_bufferSize>`, x, x, x, x
:ref:`hipsolverXgebrd <gebrd>`, x, x, x, x
.. csv-table:: Linear-systems solvers
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXpotri_bufferSize <potri_bufferSize>`, x, x, x, x
:ref:`hipsolverXpotri <potri>`, x, x, x, x
:ref:`hipsolverXpotrs_bufferSize <potrs_bufferSize>`, x, x, x, x
:ref:`hipsolverXpotrs <potrs>`, x, x, x, x
:ref:`hipsolverXpotrsBatched_bufferSize <potrs_batched_bufferSize>`, x, x, x, x
:ref:`hipsolverXpotrsBatched <potrs_batched>`, x, x, x, x
:ref:`hipsolverXgetrs_bufferSize <getrs_bufferSize>`, x, x, x, x
:ref:`hipsolverXgetrs <getrs>`, x, x, x, x
:ref:`hipsolverXXgesv_bufferSize <gesv_bufferSize>`, x, x, x, x
:ref:`hipsolverXXgesv <gesv>`, x, x, x, x
.. csv-table:: Least-square solvers
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXXgels_bufferSize <gels_bufferSize>`, x, x, x, x
:ref:`hipsolverXXgels <gels>`, x, x, x, x
.. csv-table:: Symmetric eigensolvers
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXsyevd_bufferSize <syevd_bufferSize>`, x, x, ,
:ref:`hipsolverXsyevd <syevd>`, x, x, ,
:ref:`hipsolverXsygvd_bufferSize <sygvd_bufferSize>`, x, x, ,
:ref:`hipsolverXsygvd <sygvd>`, x, x, ,
:ref:`hipsolverXheevd_bufferSize <heevd_bufferSize>`, , , x, x
:ref:`hipsolverXheevd <heevd>`, , , x, x
:ref:`hipsolverXhegvd_bufferSize <hegvd_bufferSize>`, , , x, x
:ref:`hipsolverXhegvd <hegvd>`, , , x, x
.. csv-table:: Singular value decomposition
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXgesvd_bufferSize <gesvd_bufferSize>`, x, x, x, x
:ref:`hipsolverXgesvd <gesvd>`, x, x, x, x
LAPACK-like functions
----------------------------
.. csv-table:: Symmetric eigensolvers
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverXsyevj_bufferSize <syevj_bufferSize>`, x, x, ,
:ref:`hipsolverXsyevj <syevj>`, x, x, ,
:ref:`hipsolverXsyevjBatched_bufferSize <syevj_batched_bufferSize>`, x, x, ,
:ref:`hipsolverXsyevjBatched <syevj_batched>`, x, x, ,
:ref:`hipsolverXsygvj_bufferSize <sygvj_bufferSize>`, x, x, ,
:ref:`hipsolverXsygvj <sygvj>`, x, x, ,
:ref:`hipsolverXheevj_bufferSize <heevj_bufferSize>`, , , x, x
:ref:`hipsolverXheevj <heevj>`, , , x, x
:ref:`hipsolverXheevjBatched_bufferSize <heevj_batched_bufferSize>`, , , x, x
:ref:`hipsolverXheevjBatched <heevj_batched>`, , , x, x
:ref:`hipsolverXhegvj_bufferSize <hegvj_bufferSize>`, , , x, x
:ref:`hipsolverXhegvj <hegvj>`, , , x, x
.. csv-table:: Singular value decomposition
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverDnXgesvdj_bufferSize <compat_gesvdj_bufferSize>`, x, x, x, x
:ref:`hipsolverDnXgesvdj <compat_gesvdj>`, x, x, x, x
:ref:`hipsolverDnXgesvdjBatched_bufferSize <compat_gesvdj_batched_bufferSize>`, x, x, x, x
:ref:`hipsolverDnXgesvdjBatched <compat_gesvdj_batched>`, x, x, x, x
Compatibility-only functions
====================================
The following tables summarize the wrapper functions that have not been implemented in rocSOLVER, but can be computed with an equivalent function
that uses a different algorithmic approach. These functions are provided as part of the compatibility API only, pending their implementation
in rocSOLVER.
Partial eigensolver functions
------------------------------
Partial eigensolvers have been implemented in rocSOLVER, but at present they do not use a divide-and-conquer approach.
.. csv-table:: Symmetric eigensolvers
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverDnXsyevdx_bufferSize <compat_syevdx_bufferSize>`, x, x, ,
:ref:`hipsolverDnXsyevdx <compat_syevdx>`, x, x, ,
:ref:`hipsolverDnXsygvdx_bufferSize <compat_sygvdx_bufferSize>`, x, x, ,
:ref:`hipsolverDnXsygvdx <compat_sygvdx>`, x, x, ,
:ref:`hipsolverDnXheevdx_bufferSize <compat_heevdx_bufferSize>`, , , x, x
:ref:`hipsolverDnXheevdx <compat_heevdx>`, , , x, x
:ref:`hipsolverDnXhegvdx_bufferSize <compat_hegvdx_bufferSize>`, , , x, x
:ref:`hipsolverDnXhegvdx <compat_hegvdx>`, , , x, x
Partial SVD functions
------------------------------
Partial SVD has been implemented in rocSOLVER, but at present it does not use an approximate algorithm, nor does it compute the residual norm.
.. csv-table:: Singular value decomposition
:header: "Function", "single", "double", "single complex", "double complex"
:ref:`hipsolverDnXgesvdaStridedBatched_bufferSize <compat_gesvda_strided_batched_bufferSize>`, x, x, x, x
:ref:`hipsolverDnXgesvdaStridedBatched <compat_gesvda_strided_batched>`, x, x, x, x
|
