Date: Fri, 8 May 2009 04:13:51 -0700 From: pluknet <pluknet@gmail.com> To: Scott Long <scottl@samsco.org> Cc: Ollivier Robert <roberto@keltia.freenix.fr>, freebsd-current@freebsd.org Subject: Re: Hypertherading Message-ID: <a31046fc0905080413t6f92778am1377acc06812f177@mail.gmail.com> In-Reply-To: <4A030BA1.8070709@samsco.org> References: <270637.78561.qm@web63905.mail.re1.yahoo.com> <a31046fc0905061955u4a7b5755ifbcd7bd5641cd954@mail.gmail.com> <32413E83-2059-4A47-AB45-EA7A1A509DD6@gid.co.uk> <4A030ADB.9050802@keltia.freenix.fr> <4A030BA1.8070709@samsco.org>
next in thread | previous in thread | raw e-mail | index | archive | help
2009/5/7 Scott Long <scottl@samsco.org>: > Ollivier Robert wrote: >> >> On 7/05/2009 10:17, Bob Bishop wrote: >>> >>> AFAICS the reference doesn't support that conclusion at all. >> >> Nehalem CPUs'HT feature is significantly different from the one present = in >> previous P4 CPUs. =A0Apparently, Nehalem's HT works. =A0Memory bandwidth= being >> much higher helps too. >> > > I keep here the anecdote that "it's better". =A0Is there a good reference > somewhere that describes exactly how it works? > > Scott Hi. There is a number of synthetic, low-level, and h/level application nehalem tests flowing around in Russian. Also, not far ago (31.12.2008 18:09) Intel has published the Intel Optimization Reference Manual for x32/64. (see ch. 8). It might be useful. http://download.intel.com/design/processor/manuals/248966.pdf. --=20 wbr, pluknet
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?a31046fc0905080413t6f92778am1377acc06812f177>