FAQ FAQ  Forum Search   Register Register  Login Login

Trying to grasp OC'ing...

Post Date: 2010-09-22
 Post Reply Post Reply
Author
  Topic Search Topic Search  Topic Options Topic Options
ablahblah View Drop Down
DS Veteran
DS Veteran

Email address used to purchase matched with forums account email.

Joined: 12 Jun 2009
Online Status: Offline
Posts: 2312 		  Quote ablahblah Quote  Post ReplyReply bullet Topic: Trying to grasp OC'ing...
    Posted: 22 Sep 2010 at 5:09pm
Still haven't tried my hand at it yet, just to be clear, is OC'ing essentially just tweaking around with the b-clock multiplier, and occasionally pumping up voltage to see if it'll stabilize at higher b-clocks? Still not really too clear on that. Additionally, what are the benefits of increasing other voltages such as PCI, QPI, DIMM Volt, IOH, ICH, VTT, Vcore?

Also, what are your opinions on soft-clocking? I can softclock via EVGA E-LEET, but I'm not sure how reliable that is compared to standard BIOS overclocking. Last, stress testing is usually easiest by burning it in 24-h with Prime95/LinX, but how long do you recommend to burn-in to just zero in on a prime overclock range? As in, the 24-h burn in is to completely verify if your overclock can put up with normal use, but that overclock is found by zeroing in on the hardware's maximum tolerance level with your current setup. The zeroing in testing is what I'm not sure how long to burn-in for. 15min?
R4D4RPR00F
Core i7 920 @ 3.9Ghz
Asus Sabertooth X58
EVGA GTX 570
Mushkin 6GB 1414Mhz
Back to Top
!ender_ View Drop Down
DS Veteran
DS Veteran


Joined: 24 Oct 2007
Online Status: Offline
Posts: 4219 		  Quote !ender_ Quote  Post ReplyReply bullet Posted: 22 Sep 2010 at 5:39pm
the bare bones are rasing the bclk and multi, which in turn make you change the ram ratio, which makes you change the ram timings, which makes you change the subtimings
 
think of different voltages as your wheels, each holds up its own part of the overclock as a system, examples being vcore for processor speed, dram for ram speed, and qpi/vtt for the communication between the two
 
after you get that part down, its just a race between speed and stability with temperature as the ref
 
comically enough, its a lot easier to overclock that it is to explain it, LOL
 
softclocking is pretty illadvised, its a little unpredictable and definitely not as reliable.
 
stress test for full stability is prime 95, it currently supercedes about everything for difficulty
burn testing for a dirty look at stability and temps is best done with intel burn test and or linx. the more methods you use to test, the more stable your overclock is proven to be, so the more the merrier
 
theres no right way to quick test, the only right way is through a 24 hour prime. but since that takes forever, you have to troubleshoot with other methods


Edited by !ender_ - 22 Sep 2010 at 5:42pm
Back to Top
 Post Reply Post Reply

Forum Jump Forum Permissions View Drop Down



This page was generated in 3.515625E-02 seconds.