Comments on: My time in the land of SIMD (Altivec vs SSE) http://noisymime.org/blog/2005/10/my-time-in-the-land-of-simd/ Still kickin Thu, 02 Jun 2011 16:40:26 +0000 hourly 1 http://wordpress.org/?v=3.2 By: Brian Willoughby http://noisymime.org/blog/2005/10/my-time-in-the-land-of-simd/comment-page-1/#comment-23431 Brian Willoughby Tue, 04 Aug 2009 03:38:51 +0000 http://noisymime.org/blog/?p=16#comment-23431 Did you meet all of the requirements of AltiVec, such as memory alignment, etc? You should probably share your source code if you're going to share the results, just so people can point out any mistakes that make the AltiVec look bad. SIMD is incredibly complicated. There are many easy ways to make a single mistake and lose every advantage. That just in your source code. There are also operating system concerns that you must be aware of, or your performance will drop considerably. I'm not convinced that your results mean anything, which is fine, since you say as much yourself, but I worry that this page will become a source of misinformation. One final comment: SIMD is best used for processing streams of data like audio or video. Your example of calculating pi from scratch is interesting, but it's hardly what those execution units were designed for. In other words, saying that the P4 blew away the AltiVec at something neither is primarily designed to do is not terribly accurate. Did you meet all of the requirements of AltiVec, such as memory alignment, etc? You should probably share your source code if you’re going to share the results, just so people can point out any mistakes that make the AltiVec look bad.

SIMD is incredibly complicated. There are many easy ways to make a single mistake and lose every advantage. That just in your source code. There are also operating system concerns that you must be aware of, or your performance will drop considerably. I’m not convinced that your results mean anything, which is fine, since you say as much yourself, but I worry that this page will become a source of misinformation.

One final comment: SIMD is best used for processing streams of data like audio or video. Your example of calculating pi from scratch is interesting, but it’s hardly what those execution units were designed for. In other words, saying that the P4 blew away the AltiVec at something neither is primarily designed to do is not terribly accurate.

]]> By: No more pink walls » Blog Archive » SIMD Part 2 http://noisymime.org/blog/2005/10/my-time-in-the-land-of-simd/comment-page-1/#comment-31 No more pink walls » Blog Archive » SIMD Part 2 Thu, 06 Oct 2005 20:19:58 +0000 http://noisymime.org/blog/?p=16#comment-31 [...] No more pink walls Still kickin « My time in the land of SIMD (Altivec vs SSE) SIMD Part 2 AS promised here are the resul [...] [...] No more pink walls Still kickin « My time in the land of SIMD (Altivec vs SSE) SIMD Part 2 AS promised here are the resul [...]

]]> By: tim http://noisymime.org/blog/2005/10/my-time-in-the-land-of-simd/comment-page-1/#comment-30 tim Wed, 05 Oct 2005 00:18:06 +0000 http://noisymime.org/blog/?p=16#comment-30 You can use my powermac for testing if you want. You can use my powermac for testing if you want.

]]> By: Michael http://noisymime.org/blog/2005/10/my-time-in-the-land-of-simd/comment-page-1/#comment-29 Michael Tue, 04 Oct 2005 23:41:34 +0000 http://noisymime.org/blog/?p=16#comment-29 Nice one! What is the name of that series that is used to calculate pi/4? And can I get a copy of the code to run on the Athlon64? Nice one! What is the name of that series that is used to calculate pi/4? And can I get a copy of the code to run on the Athlon64?

]]>