| Re: [eigen] Performance degradation with OpenMP |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/eigen Archives
]
- To: eigen@xxxxxxxxxxxxxxxxxxx
- Subject: Re: [eigen] Performance degradation with OpenMP
- From: Gael Guennebaud <gael.guennebaud@xxxxxxxxx>
- Date: Fri, 17 Dec 2010 20:26:28 +0100
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:mime-version:received:in-reply-to :references:from:date:message-id:subject:to:content-type; bh=8zfgUCcVn4rY5yMSBROXZ8ZoFZv7BMivA0KopHnkKX0=; b=PsMegjL4mp3kfgdQG5RJ6oA1F2ff9tfXoBgZZkamNQ6EnPJGGFK0RUA+F5VJyw0E95 QOnPAOyvKBp7Ej/qtV33pY/aZFqXSmVk5CY12QKTTGnQMAr7Ttq0f3+oOZsWSgK9qgfL MH2nm2EQe/5iGA0cD6AzbbOeuCsy31/oSSGHY=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :content-type; b=Zorc2TkBlmmyqdqJVHc+FdiRkTl/EjCN7pREG45zo0iTydSCIN1YJydwCwWmOiGV2y IzD/WnJVsCN9nY7PDhPQByPwbn099k4wI2MaBQs0x/mS6z2U12whgiB2c0J5hUzvL7cI 8G/PfSiM2Rr/O106x6YNRGEepdhQZkojJpiWw=
do you want to compute AA^T (result is 3x3) or A^TA (result is NxN)?
Because your example computes A^TA but expect a 3x3 matrix for the
result, and it fails with an assert. I changed your example to compute
AA^T (I think this is what you wanted to do), and I got exactly the
same perf. with and without OpenMP (the parallel branch is not taken).
gael
On Fri, Dec 17, 2010 at 1:18 PM, Hauke Heibel
<hauke.heibel@xxxxxxxxxxxxxx> wrote:
> Hi,
> somewhere in my code I computed A^T*A (A being 3-by-n) and I recognized a
> performance degradation of ~20% when compiling with OpenMP.
> I attached a simple test program that should be showing the issue.
> - Hauke
>