- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi All,
Why is the documentation of 15.0 version more detailed than 17.0? 17.0 doesn't even reference OpenMP on Xeon Phi as 15.0 does?
15.0: https://software.intel.com/en-us/compiler_15.0_ug_c
17.0: https://software.intel.com/en-us/intel-cplusplus-compiler-17.0-user-and-reference-guide
Also, I am on 17.x version and compiling using DeepBench [1] with KMP_PLACE_THREADS flag is working, then why 17.0 says it's depreciated? How to verify this?
icc --version
icc (ICC) 17.0.1 20161005
Copyright (C) 1985-2016 Intel Corporation. All rights reserved.
[1] https://github.com/baidu-research/DeepBench
Thanks.
- Tags:
- CC++
- Development Tools
- Intel® C++ Compiler
- Intel® Parallel Studio XE
- Intel® System Studio
- Optimization
- Parallel Computing
- Vectorization
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
hello
"OpenMP* Considerations" section for "Programming for Intel® MIC Architecture" in 15.0 https://software.intel.com/en-us/node/522515
"OpenMP* Considerations" section for "Programming for Intel® MIC Architecture" in 17.0 https://software.intel.com/en-us/node/684404
"Deprecated" word means that the feature still works but might be removed in the future releases.
hope, this helps,
--Vladimir
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
hello
"OpenMP* Considerations" section for "Programming for Intel® MIC Architecture" in 15.0 https://software.intel.com/en-us/node/522515
"OpenMP* Considerations" section for "Programming for Intel® MIC Architecture" in 17.0 https://software.intel.com/en-us/node/684404
"Deprecated" word means that the feature still works but might be removed in the future releases.
hope, this helps,
--Vladimir
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Vladimir,
Thank you. Is core and fine, the only two types of granularity KMP_AFFINITY supports? I haven't found pages where I can see types of argument an environment variable supports and technical meaning of it.
For example: Here [1], I can find details on KMP_AFFINITY, but not on technical details about the arguments this environment variable supports. Few of them are easy to understand, but some like "respect" are not.
[1] https://software.intel.com/en-us/node/522775#CC142A61-48BB-48A0-8062-92A4F266B6BE
Thanks.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
KMP_AFFINITY article is described in the "Thread Affinity Interface" chapter in the 17.0 documentation https://software.intel.com/en-us/node/684320.
I see there values and explanations with pictures
--Vladimir

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page