Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 19 Jul 2005 14:50:03 -0400
From:      "Toll, Eric" <etoll@vipstructures.com>
To:        <freebsd-amd64@freebsd.org>
Subject:   RE: better optimization flags for opteron (amd64)// -ffast-math and -O3 are not friends
Message-ID:  <9BC86C67C3AF7646B9C5382020457A944E4DA4@VIP10-WIN2K>

next in thread | raw e-mail | index | archive | help
 15834270 =3D g++ dual operon 242's @ 1.6Ghz ea. default
compile




> -----Original Message-----
> From: owner-freebsd-amd64@freebsd.org=20
> [mailto:owner-freebsd-amd64@freebsd.org] On Behalf Of=20
> Olexandr Konovalenko
> Sent: Tuesday, July 19, 2005 11:26 AM
> To: freebsd-amd64@freebsd.org
> Subject: Re: better optimization flags for opteron
(amd64)//=20
> -ffast-math and -O3 are not friends
>=20
> hi!
>=20
> I found that on my system combination of -ffast-math AND
-O3=20
> is slower than -O0 on AMD64 (interestingly it is true for=20
> Xeon EM64 as well), how can it be?

Also true for Opteron.

My Score was: 15834270 =3D g++ dual operon 242's @ 1.6 default
compile options, -O3 was 14944704

Interesting program, now *who/what* can beat the score of:
20152707 ????

(Post cpu specs please, I like to see an SMP Opteron 248
score...)



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9BC86C67C3AF7646B9C5382020457A944E4DA4>