It should be doable. Intel probably won't look too kindly on such a performance comparison..
:) The following article: http://arstechnica.com/articles/paedia/cpu/hyperthreading.ars/5 basically states that the design of HT has some shortfalls in the area of shared resources. Certain applications designed with SMP in mind (SQL SERVER!) degrade under HT situations due to the fact that there aren't actually 2 cpu's and 2 sets of cache to work with. Any server application designed for SMP would be better on a dual-core platform vs. a HT platform. Dustin Puryear wrote: >Interesting indeed. I wonder if applications that have been redesigned to >work better on multicore CPUs would perform better on Intel's HT-capable >chips, or if this is simply a defect in HT and that it won't work in many >real-world, high-load situations? (Multicore CPUs cerrtainly aren't going to >benefit every application realm either.) > >What would be interesting is to compare a HT chip with, say, a dual-core >chip for the same application. You'd have to compare chips running at >comparable clock speeds of course. Is that even doable? > >--- >Puryear Information Technology, LLC >Baton Rouge, LA * 225-706-8414 >http://www.puryear-it.com > >Author of "Best Practices for Managing Linux and UNIX Servers" >Download your free copy: >http://www.puryear-it.com/bestpractices.htm > >----- Original Message ----- >From: John Hebert >To: General at brlug.net >Sent: Saturday, November 19, 2005 9:56 AM >Subject: [brlug-general] slashdot: "HyperThreading hurts server >performance";I'm seeing the problem. Anybody else? > > >Howdy, > >My department at work runs an MS SQL Server on an Intel HyperThreaded >machine and we are seeing consistently high CPU loads lately. In fact, the >high CPU load started causing problems about a week ago with UPDATE queries >timing out for an app that I am responsible for. > >This morning, I came across the following article on Slashdot.org: > >Hyperthreading Hurts Server Performance? >"ZDNet is reporting that enabling Intel's new Hyperthreading Technology on >your servers could lead to markedly decreased performance, according to some >developers who have been looking into problems that have been occurring >since HT has been shipping automatically activated. One MS developer from >the SQL server team put it simply: 'Our customers observed very interesting >behaviour on high-end HT-enabled hardware. They noticed that in some cases >when high load is applied SQL Server CPU usage increases significantly but >SQL Server performance degrades.' Another developer, this time from Citrix, >was just as blunt. 'It's ironic. Intel had sold hyperthreading as something >that gave performance gains to heavily threaded software. SQL Server is very >thread-intensive, but it suffers. In fact, I've never seen performance >improvement on server software with hyperthreading enabled. We recommend >customers disable it.'" > >The ZDNet article is at: >http://news.zdnet.co.uk/0,39020330,39237341,00.htm > >Anybody else notice a problem with Intel HT for heavily-threaded server >applications (like databases)? I'm thinking of closely measuring various >metrics on that server then disabling HT and measuring again to compare the >results. Other ideas? > >Thanks, >John Hebert > > > > >_______________________________________________ >General mailing list >General at brlug.net >http://brlug.net/mailman/listinfo/general_brlug.net > > >_______________________________________________ >General mailing list >General at brlug.net >http://brlug.net/mailman/listinfo/general_brlug.net > > >
