This bugzilla service is closed. All entries have been migrated to https://gitlab.com/libeigen/eigen
Bug 1259 - Avoid conflicts with Intel Short Vector Math Library
Summary: Avoid conflicts with Intel Short Vector Math Library
Status: NEW
Alias: None
Product: Eigen
Classification: Unclassified
Component: Core - vectorization (show other bugs)
Version: 3.4 (development)
Hardware: x86 - general All
: Normal Optimization
Assignee: Nobody
URL:
Whiteboard:
Keywords: JuniorJob
: 1681 (view as bug list)
Depends on:
Blocks: 1687
  Show dependency tree / graph
 
Reported: 2016-07-26 09:28 UTC by Gael Guennebaud
Modified: 2019-12-04 16:01 UTC (History)
5 users (show)


Attachments
Add an attachment (proposed patch, testcase, etc.)

Description Gael Guennebaud 2016-07-26 09:28:50 UTC 
Recent Intel's compiler are able to automatically vectorize math functions such as cos, exp, etc. through calls to SVML. Apparently, SVML versions are faster than ours, and therefore disabling Eigen's vectorization might lead to faster code...

One possible strategy would be to perform ourself explicit calls to SVML in our pexp, p* functions when the user compiles with -DEIGEN_USE_SVML.
Comment 1 neumann 2019-02-13 12:47:56 UTC 
Just a note here:
Visual Studio 2019 ships with SVML intrinsics (svmlintrin.h included by immintrin.h).

Would be nice if Eigen would call those functions using visual studio.
Comment 2 Christoph Hertzberg 2019-02-15 12:47:30 UTC 
*** Bug 1681 has been marked as a duplicate of this bug. ***
Comment 3 Nobody 2019-12-04 16:01:00 UTC 
-- GitLab Migration Automatic Message --

This bug has been migrated to gitlab.com's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.com/libeigen/eigen/issues/1259.
Note You need to log in before you can comment on or make changes to this bug.