Welcome Guest ( Log In | Register )

Bump Topic Topic Closed RSS Feed
4 Pages  1 2 3 > » Bottom

Outline · [ Standard ] · Linear+

 Sandy Bridge Overclocking V1, 2500K/2600K/2700K | LGA1155 P67.Z68.Z77

views
     
cstkl1
post Jan 20 2011, 09:07 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

QUOTE(nncl @ Jan 20 2011, 08:12 PM)
guys, wanna ask something here about the memory pairing with sandy bridge cpu before i buy sandy bridge. because now still not many ram special made for sandy on sale ( except ripjaws x series, but hard to get in malaysia). so my question is will there any chance a kingston 2133c9 , dominator gt 2133c9 or dominator gt 2000c8 and clock the speed at 2200 cl7 something like the ripjaws x speed? cause i know the sandy bridge memory controller is stronger right?
*
sandybridge is weaker than a p55 based cpu

the bandwidth gain on everest is flaky
u get correct reading on sandra

everest gain has always been more cpu dependent first, and then mem controller and then mem.

sandra is more on mem and then mem controller.

so far issue on sb has been with elpida hypers trcd's and trtp's and psc elpida bbse variant ( aka the 2k c6, 2533 c9)
are having issue. this aint a bios issue just that its not conforming with intel jdec on sb IMC subtimings
cstkl1
post Jan 28 2011, 08:07 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

user posted image

user posted image
cstkl1
post Jan 28 2011, 11:26 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

Will be using M4e for the time being until i complete this new rig.

Whichever mobo that comes with a full coverage block is the preference

so far m4e vs sabertooth..seriously no diff.
just that m4e has a lot of redundant options
cstkl1
post Feb 1 2011, 03:51 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

H67 cannot overclock on multi


cstkl1
post Feb 2 2011, 10:50 AM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

nice
but isnt ure glops again bugged with the cpu pll.

i get 84 on 4.7 HT on

i found that the pll volage seriously gives inconsistent bug linx runs


cstkl1
post Feb 2 2011, 03:15 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

yeah thats the thing i also dont get with this sandybridge.

higher pll.. glops suffers
cstkl1
post Feb 2 2011, 09:19 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

QUOTE(owikh84 @ Feb 2 2011, 06:35 PM)
Maximus tends to BCLK up itself from 100 to 101.5, while other mobos is still 98.5 at default lol. I got 84GFlops when up vcore a few notches and BCLK shot up itself to 101.5 during LinX.

i dont have such a issue


Added on February 2, 2011, 6:57 pm
GFlops will either up or down. GFlops will be constant once ure at optimum vcore. The GFlops wont increase after this point. GFlops increases when BCLK is increased. But GFlops will drop when u push CPU PLL higher. Means CPU is not giving the actual performance in terms of GFlops when u alter the CPU PLL.
*
eh no

pll is based on cpu speed to dram speed.nothing to do with vcore.
u find the optimum pll for the corresponding ratio.
not to the corresponding vcore.
vcore has nothing to do with it.



cstkl1
post Feb 3 2011, 12:12 AM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

QUOTE(monsh @ Feb 2 2011, 11:02 PM)
@owikh

is that some sort of features provided by asus in m4e ? auto boost bclk on stress system ?

@cstkl1

so ur saying that vCore has nothing to do with the PLL is it ?
optimize both for best performance is it ?

does this mean that at optimum vCore and PLL the SB will perform at its best ?

but increasing vCore might reduce or improve system but resulting in not completely stable system ?

and higher PLL simply means lower gFlops ?
*
cpu pll has always been cpu frequency to dram speed related.

vcore in the sense of oscillation i guess u can call it related.

it was the same in x58. hence y u see my clocks having high glops.


i've been running at 1.5v up to 1.47 so far.

for i7 was running at 1.5 up to 4.3ghz and then for 4.4-4.5 was 1.65v. i did a test with 1.83 i think for bclk 216 but didnt yield anything
cstkl1
post Feb 5 2011, 07:29 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

linx is flawed on HT even with the new linpack libary

4.8ghz 1.41v HT on temps max 75/76C
4.8ghz 1.39v HT off temps max at 82/83C

lol

this is with my already crapped out alphacool yellowstone because of the indigo extreme.

will try with another block and hope fully with indigo extreme again.
me thinkest indigo will love this sandy bridge because it performs well in high temps only aka 75++

also hmm testing that pll overvoltage thingy.. it seems this is the main cause of the lower glops
not really pll voltage persay

This post has been edited by cstkl1: Feb 5 2011, 07:41 PM
cstkl1
post Feb 5 2011, 09:24 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

not just ht dude..

l3 cache..thats a biggie difference.
cstkl1
post Feb 5 2011, 10:15 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

bugged
y u think no cpuz lol

test 4850 2200mhz CL7 HT on 1.44v
Prime95 Blend 3250mb Max temp 82C - 2hours run
Linx with lates Linpack 3250mb Max Temp 77C - 2 hours run
4850mhz HT off 1.41v
Prime 79C 30 mins run
Linx 84C

cstkl1
post Feb 5 2011, 11:29 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

prime if u do anything less than 12 hour is just considered a memtest

finish the whole fft's

currently i dont trust any HT on test.
lol..

just running atm 4850ghz at 2200mhz CL7 HT off

will try higher when i replace my cpu block and TIM

This post has been edited by cstkl1: Feb 5 2011, 11:31 PM
cstkl1
post Feb 6 2011, 02:06 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

QUOTE(owikh84 @ Feb 6 2011, 12:13 PM)
PLL doesnt hurt my GFolps, turning off AI suite helps me in getting 84Glops on HT and 132GFlops HT off. Meanwhile i just need to enable internal PLL overvolt 51x onwards.
*
eh would really like to see that 5ghz Ht off.

curious since ure ht on temp hit 81C
cstkl1
post Feb 6 2011, 06:05 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

yeah saw that.
cstkl1
post Feb 8 2011, 01:40 AM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

Dom GT testing 103.2x47 , 2200mhz 7-10-7-28 1n at 1.625v/1.607v/1.6135v(Set/Bios Read/Dmm Read)
user posted image
cstkl1
post Feb 11 2011, 05:55 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

QUOTE(ron4 @ Feb 11 2011, 04:57 PM)
Hi im using i5 2400 and Asus P8P67 mobo. For this chipset design flaw, am i have to return back this processor and mobo to whom actually?, please guide me..

So far my rig works fine.
*
proc has no flaw

mobo distro already said u can do 1:1 in april when new chipset comes out
talk to ure retailer
if u bought from online reseller then LOL!!!
cstkl1
post Feb 12 2011, 11:06 AM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

hmm contemplating should i bother to change when already hit my target

4.85 HT off 2250 C7-10-7-28 1.65v

looking now for some rams lol that can lower that vdimm atleast to 1.625

theres only three kits hat possible me thinkest..

This post has been edited by cstkl1: Feb 12 2011, 11:08 AM
cstkl1
post Feb 12 2011, 06:39 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

Lol nope hell no for the first two

GTX4 is correct the 2533 C9
Gskill 6gb kit 2k C6 ( not the 4gb kit)
Geil Ultra 6gb 2k C6
OCZ XTE Plat 6gb 2k c6

this should be the best of the best

This post has been edited by cstkl1: Feb 12 2011, 10:11 PM
cstkl1
post Feb 15 2011, 04:05 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

running two windows atm

well my port 0,1 PCH is running C300 raid 0 on Windows Ultimate 64 SP1
and 2,3,4 is running intel x25m Raid 0 on Xp Proffessional 64bit SP3

so far no issue

cstkl1
post Feb 23 2011, 06:17 PM

Look at all my stars!!
Group Icon
Elite
6,799 posts

Joined: Jan 2003

Lol farny
4sticks are dual channel.
And increase in vdimm is because of loss of vdimm in traceline and hence inconsistent termination voltage

This post has been edited by cstkl1: Feb 23 2011, 06:19 PM

4 Pages  1 2 3 > » Top
Topic ClosedOptions
 

Change to:
| Lo-Fi Version
0.0447sec    0.21    7 queries    GZIP Disabled
Time is now: 4th December 2025 - 07:55 PM