Thread: AMD 12 cores
View Single Post
  #38 (permalink)  
Old April 8, 2010, 02:17 PM
m1dget's Avatar
m1dget m1dget is offline
Allstar
 
Join Date: Nov 2008
Location: Terrebonne, Qc
Posts: 707

My System Specs

Default

Quote:
Originally Posted by Zero82z View Post
So what if it's the same number of threads? You can't compare HT logical cores to real cores in terms of how much can be done with each thread. They're still running on half as many cores as the AMD system.
Well yes and no. If you want to compare the brute speed of both cpu, you don't have much choice here, you have to put them head to head like in your link.

Quote:
Originally Posted by Zero82z View Post
Keep in mind that the AMD system was running quad-channel RAM and the Intel system was running triple-channel. As for the Beckton with its quad-channel memory controller, as stated in the review it looks like there were some issues with the testing methods, because the results just don't make sense.
It depends on what went wrong... IMO doing the cinebench and graphical bench test is plain retarded because most of these chips are going to end up either in servers doing 'server stuff' or doing calculation and most likely will not be used in desktops or even maybe workstations.
I might be wrong on that though because if I was anyone who had to do heavy threaded applications I would run AWAY from x86-64 and go with the ultrasparc architecture (that's -heck- of a nice one).

Quote:
Originally Posted by Zero82z View Post
Applications with near perfect scaling should theoretically perform much better on the Magny-Cours system because it has more real cores.
Yes... that's granted.
__________________
"NIX is a classic example of security through obscurity because there is no real monetary reward for crackers and hackers to break Linix" -AkG
Reply With Quote