From owner-freebsd-amd64@FreeBSD.ORG Wed Feb 29 08:40:35 2012 Return-Path: Delivered-To: freebsd-amd64@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 8791C106564A for ; Wed, 29 Feb 2012 08:40:35 +0000 (UTC) (envelope-from brde@optusnet.com.au) Received: from mail07.syd.optusnet.com.au (mail07.syd.optusnet.com.au [211.29.132.188]) by mx1.freebsd.org (Postfix) with ESMTP id EE9518FC0C for ; Wed, 29 Feb 2012 08:40:34 +0000 (UTC) Received: from c211-30-171-136.carlnfd1.nsw.optusnet.com.au (c211-30-171-136.carlnfd1.nsw.optusnet.com.au [211.30.171.136]) by mail07.syd.optusnet.com.au (8.13.1/8.13.1) with ESMTP id q1T8eWqx002796 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Wed, 29 Feb 2012 19:40:32 +1100 Date: Wed, 29 Feb 2012 19:40:32 +1100 (EST) From: Bruce Evans X-X-Sender: bde@besplex.bde.org To: "Thomas D. Dean" In-Reply-To: <4F4DDCE7.9000008@speakeasy.org> Message-ID: <20120229192417.U3167@besplex.bde.org> References: <4F3EA37F.9010207@speakeasy.org> <4F3EC0B4.6050107@speakeasy.org> <4F4DA398.6070703@speakeasy.org> <20120229161408.G2514@besplex.bde.org> <4F4DDCE7.9000008@speakeasy.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: freebsd-amd64@freebsd.org Subject: Re: Gcc46 and 128 Bit Floating Point X-BeenThere: freebsd-amd64@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Porting FreeBSD to the AMD64 platform List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 29 Feb 2012 08:40:35 -0000 On Wed, 29 Feb 2012, Thomas D. Dean wrote: > On 02/28/12 22:03, Bruce Evans wrote: >> >> But why would you want it? It is essentially unusable on sparc64, >> since it is several thousand times slower than 80-bit floating point >> on i386. At equal CPU clock speeds, it is only about 1000 times slower. >> ... > > I have an application that takes 10 days to run on a 4.16GHz Core-i7 3930K. > No output until it finishes. Look elsewhere :-). 1000 times slower than that would be bad :-). > When I first started looking at this, I naively thought the 80-bit FPU floats > were scaled to 128-bits. Would be nice... > > The application uses libgmp, but, about 1/2 to 2/3 of the work will fit in a > 128-bit float. libgmp is probably slow for the same reasons that multi-word software FP is slow -- it has to do similar things to combine the words. But it is much older so it it likely to be smarter about combining the words, and might already know how to use vectors on AVX. Combining the words seems to be much harder with floating point words, so you probably don't want floating point if you only need large integers. Bruce