sound like nb or mem
run windows diagnostic memory test or memtest86

orthos doesnt stress mem/nb
so many times ppl even pass 12 hours of orthos blend
but will fail even playing a simple video

so use occt.

eh
y not do a loop test on crysis say 99 times?? for cpu benchmark
that is much better test

one more to the ignore list for supporting ppl who post garbage and cant back them up

This post has been edited by cstkl1: May 7 2008, 07:41 PM

bro u know right

tRAS 2-9 around the same read result

so 4 .. means nothing.. and actually causes more instability

and tRFC 20-25 u get higher reads..

what sticks are those.. kindda bad at 2.25v

This post has been edited by cstkl1: May 8 2008, 10:28 AM

dominator??

eh..
promos??

so much vdimm??

if ure free post a everest bandwidth..
ty


This post has been edited by cstkl1: May 8 2008, 06:28 PM

eh ass soon as u can ??

or as late as u can??

kekek want to see also how this board preforms..

hint hint p5q deluxe..

its the loadline thingy
not epu thingy
epu is software based

well yeah..

it started with the maximus and rampage

on perfect overclock setting.. will pass every test on load but once in a while on idle.. comp will bsod etc

lowyatians...
a year ago when i stressed the importance of gtl..
everybody was saying not important..
even this idiot who was a elite so called was saying that
and now they are saying gtl but dont understand it..

theres no sweet spot etc and gtl's are just one value which u can find it using a vmm.. anything else is just unstable values or u just overvolted ure proc a wee bit thats about it...

all this flukish gtl's that everybody is guessing is just to lower the nb/vtt voltage to make ure nb more stable.. thats it...

not much for vcore for dualies.. more for quaddies still.


This post has been edited by cstkl1: Oct 19 2008, 09:35 AM

cryptik .. sish..

wrote that formula in the maximus 2 thread like a few months ago...

but he still doesnt get it..... gtl is just fine tuning for the vNB..
the main reason is this.. when u ure multiplier and bus to a certain vcore...
the previous FSB voltage NB voltage that works for the previous bus wont work..
reason is look at the gtl graph in www.edgeofstability.com

example for 533fsb i needed 1.16vtt, 1.32vnb gtl 0.63.. but its good up to 4.25ghz at 1.32v vcore..
because of the fact 4.5ghz i need more vcore the previous vtt voltage / vnb wont work..
even if i pum to 1.7v on vcore it wont work..i mean it will boot but wont pass any stress test to the cpu...

this is where gtl really comes in play..
its to maintain the same ratios or support resistance level ( hahahaha.. cant beleive i am actually using this word in overclocking) but the same GTL values..

i hope this explains a bit.

same thing for quads who require high vcore but not really high vNB.

cryptik wrote wrongly and took it out of text..
i wrote that just for one guy system.. not a general rule..and he is making a lot of general rules..
dont want u guys to fall into that.

http://www.xtremesystems.org/forums/showpo...9&postcount=333\

another example which all of u i think must have not realize something..
previously when u ran orthos at a certain vcore for high mHZ speed u suddenly hit a brickwall that the vcore needs to be raised exponentially.
1.so ure answer is the proc scales well to that vcore... or 2. the most commonly answer is ure proc is too hot.. ( but already disapproved when u ran linpack to 80C.. )

what if i was to tell u its not that..
this is where vid numbers really come into play...
low vid procs has neeeds lower vcore to run.. but because of this it wont suffer easily like the above issue with vNB and vFSB.
but when u look at the speed to speed of a higher vid.. the heat output for both proc and different vcore is same.
but the difference is to go higher mhz he doesnt have to tax the mobo a lot or adjust a lot..

this is the difference..
and worst still each mobo i/o.. crossovers are different etc...
so yes for a certain speed its acheivable on all the mobos but for higher speed stability.. some mobo out there might suprise u
but it only might work for ure proc.. not a general rule to all procs..

This post has been edited by cstkl1: Oct 22 2008, 04:29 PM

dude... that thing that i posted and also crptik was only for asus mobos..

its the same with the striker 2 formula/maximus 2 formula/rampage extreme/striker 2 extreme..

but in the end for more accuracy use a vmm..
again gtl i only needed example for quadcores at high vcores
or dual cores at high mhz with high vcores...

when i say high i mean like 1.4v example depending on ure nb voltages to ure fsb voltages ratio..

thats when it really comes into play...

ure xfx 780 board.. god knows why u bought that board... works differently...and doesnt clock that well ..

well all i can tell u guys

get a good mobo with a lot of good voltage adjustment features..

aka rog's... dfi's
get a VMM ..
test and document.... and u will understand ...

sorry dudes no shortcut on this..
hmm i wonder with i7 having a integrated mem controller.. gtl shouldnt be a factor anymore...

well its a damn good board..
price.. quality .. heck i still love the manual.. this board has been really well designed ...
but again its a mid range board...

so as with anything.. theres just so much it can do.. and also i remind u it depends on the proc/ure ability etc.. how good ure rams are binned etc..
so far its the only board i actually managed to get 566fsb stable..
really trying hard with the m2f.. its possible.. also.. just need a little more work..

so the question is vague dude..


secret of the acients.. want to test ure GTL/FSB Voltage...

try orthos priority 10 custom MEM at max running fft 4096...for C2d 6mb.for around 1 hour.
nowadays very easy to test stability

u want to pass any crap..
try linpack 64 50 runs at max mem in diagnostic mode..
the previous orthos setting
and memtest windows in diagnostic mode.. must load atleast 95% of ure available ram.. for 4gb 100% would do .. cause its nearly 1 hour for that.. and 2gb kits atleast 200%..
total time take to run all this test

3-4 hours only...
and it will be superstable and then u can do 100-200 linpack64 for ure own assurance.

This post has been edited by cstkl1: Oct 26 2008, 06:24 PM

dude ure proc is unstable dude...go and see ure linpack..


and run in diagnostic mode so can test all of ure rams to atleast 95percent physical mem..
for 2gb kit

no idea y u so happy with that fact..

so far tested min needed to test linpack 64 is around 1.7gbto 1.8gb... to ensure stability.. for vcore..

---

Added on October 31, 2008, 12:40 ambtw i have tested the tjmax for e8600 E0 on linpack..
its 95C..

once u touch that temp.. shutdown..
tested for 20 mins..

This post has been edited by cstkl1: Oct 31 2008, 12:41 AM

dude ures cant even pass 5 loops on less ram..

what u think is going to happen with more ram ( longer period)
64 bit is the best way to test...

32bit is like half testing... somewhere down the road.. some crash is going to happen in some long period of stressing some component... and u will never know...and ure going to be b****ing that linpack /orthos/prime doesnt mean stable..


go to start
run
type msconfig and press ok
choose diagnostic mode...
press ok...
and restart..

2gb is okay.. cause in diagnostic mode u will still get about 1.7-1.8gb worth of ram free to test...

anyway this is afar as i am gonna tell on this..
use google..

exact same thing when Nasa Started giving out Contracts to the lowest bidder..

oh wait.. didnt a few spaceshuttles/ spaces launches didnt even make it to the orbit...

my bad.

back to ignorelist u go.

This post has been edited by cstkl1: Oct 31 2008, 02:12 AM

has anybody even seen IBT 4.5ghz even 5-10 passes... posted anywhere in the world...???

hint hint its not a vcore issue that will make it fail..from all the testing i have done with multiple mobo's

---

Added on November 6, 2008, 11:30 amSome E8600 E0 Power...

Bye bye to all my ddr2 boards.. the last of them..
sus LCD Poster reading for Load
Maximus 2 Formula Loadline Enabled

VID 1.215 1.382v - GTL CPU 0/1 - 0.670/0.667


VID 1.23 1.262v - GTL CPU 0/1 - 0.635/0.667


VID 1.25 1.308v - GTL CPU 0/1 - 0.640

---

Added on November 6, 2008, 11:32 amPS the best one was the last one...
high vcore but cooler and higher bus speed easier...
max was 620 on maximus on 2 sticks..
the rest barely did 600

and for the 2nd one when i sold it.. there was a idiot in lyn who said all e8600 can do that.. what a stupid fella.

personally i think all the vid 1.25 are better.


This post has been edited by cstkl1: Nov 6 2008, 11:51 AM