You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: CHANGELOG.md
+17-22Lines changed: 17 additions & 22 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -47,7 +47,7 @@ Full documentation for ROCm Compute Profiler is available at [https://rocm.docs.
47
47
* Remove aggregated analysis view from TUI mode
48
48
49
49
50
-
## ROCm Compute Profiler 3.2.1 for ROCm 7.0.0
50
+
## ROCm Compute Profiler 3.2.3 for ROCm 7.0.0
51
51
52
52
### Added
53
53
@@ -96,7 +96,7 @@ Full documentation for ROCm Compute Profiler is available at [https://rocm.docs.
96
96
97
97
* Roofline support for RHEL 10 OS.
98
98
99
-
* FP4 and FP6 data types have been added for roofline profiling on AMD Instinct MI350 series.
99
+
*`FP4` and `FP6` data types have been added for roofline profiling on AMD Instinct MI350 series.
100
100
101
101
#### rocprofv3 support
102
102
@@ -114,17 +114,8 @@ Full documentation for ROCm Compute Profiler is available at [https://rocm.docs.
114
114
*``-b`` option in profile mode also accepts hardware IP block for filtering; however, this filter support will be deprecated soon.
115
115
*``--list-metrics`` option added in profile mode to list possible metric id(s), similar to analyze mode.
116
116
117
-
* Interface to ROCprofiler-SDK.
118
-
* Setting the environment variable ``ROCPROF=rocprofiler-sdk`` will use ROCprofiler-SDK C++ library instead of ``rocprofv3`` python script.
119
-
* Add --rocprofiler-sdk-library-path runtime option to choose the path to rocprofiler-sdk library to be used
120
-
* Using rocprof v1 / v2 / v3 interfaces will trigger a deprecation warning to use rocprofiler-sdk interface
121
-
122
117
* Support MEM chart on CLI (single run)
123
118
124
-
* Deprecation warning for MongoDB database update mode.
125
-
126
-
* Deprecation warning for ``rocm-smi``
127
-
128
119
*``--specs-correction`` option to provide missing system specifications for analysis.
129
120
130
121
### Changed
@@ -136,7 +127,12 @@ Full documentation for ROCm Compute Profiler is available at [https://rocm.docs.
136
127
* Changed the condition when Roofline PDFs are generated during general profiling and ``--roof-only`` profiling (skip only when ``--no-roof`` option is present).
137
128
* Updated Roofline binaries:
138
129
* Rebuild using latest ROCm stack
139
-
* Minimum OS distribution support minimum for roofline feature is now Ubuntu 22.04, RHEL 9, and SLES15 SP6.
130
+
* Minimum OS distribution support minimum for roofline feature is now Ubuntu 22.04, RHEL 8, and SLES15 SP6.
131
+
132
+
### Removed
133
+
134
+
* Roofline support for Ubuntu 20.04 and SLES below 15.6
135
+
* Removed support for AMD Instinct MI50 and MI60.
140
136
141
137
### Optimized
142
138
@@ -146,7 +142,10 @@ Full documentation for ROCm Compute Profiler is available at [https://rocm.docs.
146
142
147
143
* Fixed kernel name and kernel dispatch filtering when using ``rocprofv3``.
148
144
* Fixed an issue of TCC channel counters collection in ``rocprofv3``.
149
-
* Fixed peak FLOPS of F8, I8, F16, and BF16 on AMD Instinct MI 300.
145
+
* Fixed peak FLOPS of `F8`, `I8`, `F16`, and `BF16` on AMD Instinct MI300.
146
+
* Fixed not detecting memory clock issue when using amd-smi
147
+
* Fixed standalone GUI crashing
148
+
* Fixed L2 read/write/atomic bandwidths on AMD Instinct MI350 series.
150
149
151
150
### Known issues
152
151
@@ -165,19 +164,15 @@ Full documentation for ROCm Compute Profiler is available at [https://rocm.docs.
165
164
166
165
* Memory chart on ROCm Compute Profiler CLI might look corrupted if the CLI width is too narrow.
167
166
168
-
### Removed
169
-
170
-
* Roofline support for Ubuntu 20.04 and SLES below 15.6
171
-
* Removed support for AMD Instinct MI50 and MI60.
167
+
* Roofline feature is currently not functional on Azure Linux 3.0 and Debian 12.
172
168
173
169
### Upcoming changes
174
170
175
-
*``rocprof v1/v2/v3`` interfaces will be removed in favor of the ROCprofiler-SDK interface, which directly accesses ``rocprofv3`` C++ tool.
176
-
* To use ROCprofiler-SDK interface, set environment variable `ROCPROF=rocprofiler-sdk` and optionally provide profile mode option ``--rocprofiler-sdk-library-path /path/to/librocprofiler-sdk.so``
171
+
*``rocprof v1/v2/v3`` interfaces will be removed in favor of the ROCprofiler-SDK interface, which directly accesses ``rocprofv3`` C++ tool. Using ``rocprof v1/v2/v3`` interfaces will trigger a deprecation warning.
172
+
* To use ROCprofiler-SDK interface, set environment variable `ROCPROF=rocprofiler-sdk` and optionally provide profile mode option ``--rocprofiler-sdk-library-path /path/to/librocprofiler-sdk.so``. Add ``--rocprofiler-sdk-library-path`` runtime option to choose the path to ROCprofiler-SDK library to be used.
177
173
* Hardware IP block based filtering using ``-b`` option in profile mode will be removed in favor of analysis report block based filtering using ``-b`` option in profile mode.
178
-
* Using rocprof v1 / v2 / v3 interfaces will trigger a deprecation warning to use rocprofiler-sdk interface
179
-
* MongoDB database support will be removed.
180
-
* Usage of ``rocm-smi`` will be removed in favor of ``amd-smi``.
174
+
* MongoDB database support will be removed, and a deprecation warning has been added to the application interface.
175
+
* Usage of ``rocm-smi`` is deprecated in favor of ``amd-smi``, and a deprecation warning has been added to the application interface.
Copy file name to clipboardExpand all lines: docs/how-to/profile/mode.rst
-1Lines changed: 0 additions & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -509,7 +509,6 @@ successfully.
509
509
.. note::
510
510
511
511
* ROCm Compute Profiler currently captures roofline profiling for all data types, and you can reduce the clutter in the PDF outputs by filtering the data type(s). Selecting multiple data types will overlay the results into the same PDF. To generate results in separate PDFs for each data type from the same workload run, you can re-run the profiling command with each data type as long as the ``roofline.csv`` file still exists in the workload folder.
512
-
* Roofline feature is currently not enabled on AMD Instinct MI350.
513
512
514
513
The following image is a sample ``empirRoof_gpu-0_FP32.pdf`` roofline
0 commit comments