From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.1.3 (2006-06-01) on yquem.inria.fr X-Spam-Level: X-Spam-Status: No, score=0.0 required=5.0 tests=AWL autolearn=disabled version=3.1.3 Received: from mail1-relais-roc.national.inria.fr (mail1-relais-roc.national.inria.fr [192.134.164.82]) by yquem.inria.fr (Postfix) with ESMTP id DA258BBAF for ; Fri, 11 Jul 2008 20:14:42 +0200 (CEST) X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AsQBAAJAd0g+FFrDiGdsb2JhbACSKAEBAQ8gnR8 X-IronPort-AV: E=Sophos;i="4.30,346,1212357600"; d="scan'208";a="15015895" Received: from dns.vtab.com (HELO oden.vtab.com) ([62.20.90.195]) by mail1-smtp-roc.national.inria.fr with ESMTP/TLS/DHE-RSA-AES256-SHA; 11 Jul 2008 20:14:42 +0200 Received: from oden.vtab.com (oden.vtab.com [127.0.0.1]) by oden.vtab.com (Postfix) with ESMTP id 55BFF26EF1C; Fri, 11 Jul 2008 20:14:41 +0200 (CEST) Received: from kicki.hq.vtech (kicki.hq.vtech [10.0.0.211]) by oden.vtab.com (Postfix) with ESMTP id 4301E26EF09; Fri, 11 Jul 2008 20:14:41 +0200 (CEST) Received: by kicki.hq.vtech (Postfix, from userid 7557) id 0DE498337C; Fri, 11 Jul 2008 20:14:40 +0200 (CEST) From: "=?utf-8?b?TWF0dGlhcyBFbmdkZWfDpQ==?= =?utf-8?b?cmQ=?=" To: Xavier.Leroy@inria.fr Cc: jhc0033@gmail.com, caml-list@yquem.inria.fr In-reply-to: <487768E2.6000108@inria.fr> (message from Xavier Leroy on Fri, 11 Jul 2008 16:06:26 +0200) Subject: Re: [Caml-list] thousands of CPU cores References: <487768E2.6000108@inria.fr> Message-Id: <20080711181440.0DE498337C@kicki.hq.vtech> Date: Fri, 11 Jul 2008 20:14:40 +0200 (CEST) X-Virus-Scanned: ClamAV using ClamSMTP X-Spam: no; 0.00; mattias:01 mattias:01 caml-list:01 computing:05 shared:06 shared:06 needing:07 thousands:91 useful:09 i'm:09 memory:09 memory:09 practical:09 machine:09 actually:10 >[...] There are good reasons to think that the >illusion of shared memory cannot be maintained in the presence of >hundreds of computing elements, even using cc-NUMA techniques >(i.e. hardware emulation of shared memory on top of high-speed >point-to-point links). I'm not arguing any of your points but just note that larger NUMA machines than that are available and sometimes practical - SGI Altix go up to 1024 cores with a single system image. (To answer Richard Jones's question, I know Bea have tested their JVM on such a machine but I have no idea whether it turned out to be useful. I doubt there are many Java applications actually needing such a wide JVM.)