From owner-sc22wg5+sc22wg5-dom8=www.open-std.org@open-std.org Fri Apr 5 23:50:43 2013 Return-Path: X-Original-To: sc22wg5-dom8 Delivered-To: sc22wg5-dom8@www.open-std.org Received: by www.open-std.org (Postfix, from userid 521) id 6FD4A356D36; Fri, 5 Apr 2013 23:50:43 +0200 (CEST) Delivered-To: sc22wg5@open-std.org Received: from ppsw-50.csi.cam.ac.uk (ppsw-50.csi.cam.ac.uk [131.111.8.150]) by www.open-std.org (Postfix) with ESMTP id 5C2D6356D36 for ; Fri, 5 Apr 2013 23:50:41 +0200 (CEST) X-Cam-AntiVirus: no malware found X-Cam-SpamDetails: not scanned X-Cam-ScannerInfo: http://www.ucs.cam.ac.uk/email/scanner/ Received: from hermes-1.csi.cam.ac.uk ([131.111.8.51]:34185) by ppsw-50.csi.cam.ac.uk (smtp.hermes.cam.ac.uk [131.111.8.157]:25) with esmtpa (EXTERNAL:nmm1) id 1UOEWu-0004vy-rf (Exim 4.72) (return-path ); Fri, 05 Apr 2013 22:50:40 +0100 Received: from prayer by hermes-1.csi.cam.ac.uk (hermes.cam.ac.uk) with local (PRAYER:nmm1) id 1UOEWu-0003OM-K9 (Exim 4.72) (return-path ); Fri, 05 Apr 2013 22:50:40 +0100 Received: from [87.112.94.53] by webmail.hermes.cam.ac.uk with HTTP (Prayer-1.3.5); 05 Apr 2013 22:50:40 +0100 Date: 05 Apr 2013 22:50:40 +0100 From: "N.M. Maclaren" To: sc22wg5 Subject: Re: [ukfortran] (SC22WG5.4968) (j3.2006) AW: AW: WG5 ballot on first draft TS 18508, Additional Parallel Features in Fortran (Update) Message-ID: In-Reply-To: <20130405203929.732E2356CA2@www.open-std.org> References: <20130329104945.2C28A356BB3@www.open-std.org> <20130329232930.A368A356DC2@www.open-std.org> <20130330132208.D7468356DDA@www.open-std.org> <20130330141442.761E2356D8C@www.open-std.org> <20130405203929.732E2356CA2@www.open-std.org> X-Mailer: Prayer v1.3.5 Mime-Version: 1.0 Content-Type: text/plain; format=flowed; charset=ISO-8859-1 Sender: owner-sc22wg5@open-std.org Precedence: bulk On Apr 5 2013, Bill Long wrote: >> >>> I must however admit that a significant (if not the bulk) part of the >>> FCA optimization is concerned with topology awareness, so will also be >>> applicable to the general reduction facility. >> >> How much of that gets through the driver layer is less clear. When I >> last looked, OpenIB didn't support RDMA, which makes a lot of the fancy >> MPI one-sided and PGAS stuff irrelevant. > >Not sure about OpenIB, but the Mellanox IB hardware certainly supports >RDMA. Oh, yes, but you need a complete path from the interface card to the application before RDMA is of much use, and one broken link is enough to make it useless. The problem is that RDMA is very hard to implement in a semi-portable interface, because of its horrible interactions with the kernel's memory management and protection. Regards, Nick Maclaren.