- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello, i have read Intel Cilk Plus documentation, but there is one issue with CEAN (C/C++ extended array notation) i'm confused about. Does it use thread level paralellism or only SIMD level parallelism? Or more specifically, an i use it to just vectorise my code without adding any multithreading?
1 Solution
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello Sergey,
As far as I know, Array notation only adds SIMD level parallelism.
Thanks,
Balaji V. Iyer.
As far as I know, Array notation only adds SIMD level parallelism.
Thanks,
Balaji V. Iyer.
Link Copied
2 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello Sergey,
As far as I know, Array notation only adds SIMD level parallelism.
Thanks,
Balaji V. Iyer.
As far as I know, Array notation only adds SIMD level parallelism.
Thanks,
Balaji V. Iyer.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Array notation is useful in order to employ Cilk+ templates:
sum = __sec_reduce_add(a[:] > 0 ? a[:] : 0); // sum of positive elements
Apparently this doesn't involve either vectorization or threading; it's remarkably efficient under those limitations.
sum = __sec_reduce_add(a[:] > 0 ? a[:] : 0); // sum of positive elements
Apparently this doesn't involve either vectorization or threading; it's remarkably efficient under those limitations.

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