Matrix Multiplication with C++ AMP (Channel 9)

Matrix Multiplication with C++ AMP

Posted: Jul 03, 2012 at 5:38 PM

By: Daniel Moth

(5)

23,165 Views

8 Comments

Video format

Option selected may change based on video formats available and browser capability.

Auto

Use our suggested format for this video.

Smooth Streaming

Adaptive bitrate streaming - little buffering but lower video quality at times of low bandwidth. Silverlight 5 required

Progressive

Traditional buffered video format. Silverlight 5 required

HTML5

For browsers that support H.264 MP4 or WebM video playback such as:
IE9+, Firefox 4+, Chrome 5+, or Safari 4+.

Flash

For browsers that have the Adobe Flash Player plugin installed.

Hi, I am Daniel Moth Smiley

This screencast assumes knowledge of the C++ AMP API, i.e. the simple model and the tiled model, and here we look at a common algorithm (matrix multiplication) and convert it from serial on the CPU, to run on the GPU, first with the simple model and then tiled.

To learn more please visit the C++ AMP blog, and we encourage C++ AMP questions in the Parallel Computing in C++ and Native Code MSDN forum.

Follow the Discussion

Follow
Oops, something didn't work.

Try again?

Sign In to be begin to follow these comments

What does this mean?
Following an item on Channel 9 allows you to watch for new content and comments that you are interested in. You need to be signed in to Channel 9 to use this feature.

Getting "follow" information

Stop following these comments

Start following these comments

What does this mean?
Following an item on Channel 9 allows you to watch for new content and comments that you are interested in and view them all on your notifications page.

Removing your "follow"

Setting up your "follow"

Did you know you can
sign up for email notifications?
- Subscribe to this comments

Aiboy

Jul 06, 2012 at 10:00 AM

Awesome Tnx a lot for this videocast...
Spetum

Jul 06, 2012 at 11:34 AM

not niche!!
Thank you for this video lecture.
Elad

Jul 07, 2012 at 12:34 AM

Great video!
It was very interesting to watch.

You should have showed the time it takes for each version at the end....
Now I have to write the code and try it. :)

Thanks!
Daniel Moth

Jul 10, 2012 at 7:13 PM

@Spetum: @Aiboy: Thanks, glad you enjoyed it.
@Elad:Thanks, glad you found it interesting. Yes I leave the time measurement to you so you can explore the benefits on your specific hardware
martin w

Aug 31, 2012 at 6:43 PM

very nice video! i have some spare time these days so i'm giving AMP a spin.
one question though: i'm not sure if i understood this correctly but does your tiled version require the input to be of even length (because of TS = 2)? I'm relatively new to AMP. comparing with what i know from using sse intrinsics for example (not really comparable, i know, but some concepts are at least similar) i'd expect some code for dealing with odd length vectors i.e. padding or computing the rest in a non-tiled way (painful). did i miss something here? im sure i did..

cheers!
martin
Daniel Moth

Sep 02, 2012 at 2:49 PM

@martin w: Thanks, glad you enjoyed it!
The answer to your question is in this blog post and the links it points to: http://blogs.msdn.com/b/nativeconcurrency/archive/2012/02/26/divisibility-requirement-of-tiled-extent-in-c-amp.aspx
If you still have questions after reading those, please post to our forum: http://social.msdn.microsoft.com/Forums/en-US/parallelcppnative/threads
J777

Sep 24, 2012 at 7:26 AM

Many thanks for the awesome tutorials :)
faisalharoon

Oct 18, 2012 at 11:46 AM

very nice video! i have some spare time these days so i'm giving AMP a spin.