From mboxrd@z Thu Jan 1 00:00:00 1970 Received: (from majordomo@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id IAA04385; Wed, 22 Jan 2003 08:22:18 +0100 (MET) X-Authentication-Warning: pauillac.inria.fr: majordomo set sender to owner-caml-list@pauillac.inria.fr using -f Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39]) by pauillac.inria.fr (8.7.6/8.7.3) with ESMTP id IAA03854 for ; Wed, 22 Jan 2003 08:22:17 +0100 (MET) Received: from dux.ru (ns.dux.ru [193.125.210.65]) by concorde.inria.fr (8.11.1/8.11.1) with ESMTP id h0M7MGr22267 for ; Wed, 22 Jan 2003 08:22:16 +0100 (MET) Received: from dux.ru (localhost [127.0.0.1]) by dux.ru (8.12.6/8.12.6) with ESMTP id h0M7QB1c010212 for ; Wed, 22 Jan 2003 10:26:11 +0300 (MSK) (envelope-from snob@snob.spb.ru) Received: (from uucp@localhost) by dux.ru (8.12.6/8.12.1/Submit) with UUCP id h0M7QBg1010211 for caml-list@inria.fr; Wed, 22 Jan 2003 10:26:11 +0300 (MSK) (envelope-from snob@snob.spb.ru) Received: (qmail 1446 invoked by uid 0); 21 Jan 2003 22:37:44 -0000 Received: from unknown (HELO athlon.snob.spb.ru) (unknown) by unknown with SMTP; 21 Jan 2003 22:37:44 -0000 Content-Type: text/plain; charset="iso-8859-1" From: Nickolay Semyonov-Kolchin To: David Chase , caml-list@inria.fr Subject: Re: Coyote Gulch test in Caml (was Re: [Caml-list] speed ) Date: Wed, 22 Jan 2003 01:26:33 +0500 User-Agent: KMail/1.4.3 References: <5.2.0.9.0.20030120161510.066e4598@pop.theWorld.com> <5.2.0.9.0.20030121080628.068047f0@pop.theWorld.com> In-Reply-To: <5.2.0.9.0.20030121080628.068047f0@pop.theWorld.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Message-Id: <200301220126.33978.snob@snob.spb.ru> Sender: owner-caml-list@pauillac.inria.fr Precedence: bulk On Tuesday 21 January 2003 18:09, David Chase wrote: > At 02:39 AM 1/21/2003 +0500, Nickolay Semyonov-Kolchin wrote: > >Speed and accuracy are different things. Matlab class software need > > accuracy, most computer games need speed. This is the reason of > > "-ffast-math" key in gcc. Ocaml lacks such key, and always produce > > ineffecient floating-point code. > > But how much accuracy do computer games need? First-class implementati= ons > of sin/cos in software are quite fast, indeed faster (in certain > non-trivial ranges) than the hardware itself. If it happens that you c= ould > determine how *little* accuracy you actually need, it could go faster y= et > :-). > That depends. "Computer games" was just an example of software that don't= =20 require great accuracy. Nickolay ------------------- To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/ Beginner's list: http://groups.yahoo.com/group/ocaml_beginners