Jump to content


AMD "behind" Intel? think again...


  • Please log in to reply
16 replies to this topic

#16 m.oreilly

m.oreilly

    rog'er wilco

  • Admin
  • PipPipPipPipPipPip
  • 8,847 posts
  • Country:lower uncton

Posted 10 December 2007 - 03:07 AM

after two days of pricing out an intel xeon 5400 series dual quad rig (with seaburg chipset mobo), i stumbled on this thread over at 2cpu, containing this review . i had originally started another thread there re: the new intel server boards, but was glad i decided to browse a bit more. after all the ultra bad press amd is getting atm (granted, they are having problems with their quad stuff), their 2.5gig 'messed up' quad cpu did a great showing against the latest intel 3gig 45nano, actually beating down the 5300 xeon series quads, also at 3gigs. remember, this is server/workstation stuff. i think this quote from 'zane' sheds some light on earlier amd design, and what purchasing multi core systems to use with multi threaded apps is all about:

Quote

I tested a memory intensive code on E5345 and Opteron 270. E5345 is about 60% faster than Opteron 270 with one thread. But the speed up of E5345 is about 2.5 with eight threads, while the speed up of Opteron 270 is about 6.5 with eight threads. So the Opteron 270 is about 60% (6.5/(2.5*1.6)=1.625) faster than E5345 with eight threads. E54xx has 40%~60% more bandwidth compared to E53xx. I estimate that the Xeon E5420 and Opteron 2347 should have similar performance(for the code). Both have similar price. It will not be wrong for me to get either one.

i just might go for an intel xeon rig this time, as i have never owned one, and it will give me some street cred if i ever decided to do some intel bashing :likewinmandr2:

#17 Guest_scaramonga_*

Guest_scaramonga_*
  • Guests

Posted 10 December 2007 - 04:19 AM

Oh how the mighty fall :likewinmandr2: :cc_surrender:




1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users