FSX doesn't seem to like my new clock

[vgbaron]

same

[NickN]

Let me tell you what is really screwed up... Mushkin setting a 70mv GTL would suggest their sticks are using the .63x multiplier and not the VTT mv formula BIOS engineering uses

I really need to talk to a mush engineer to find out what the skinny is on this because based on the VTT (FSB Termination Voltage) to GTL formula

Vtt x 0.667 = Vtt x 0.635 + Y

that value should be @ a 1.1 FSB TERM VOLT = -35mv

[vgbaron]

understood!

[NickN]

What I dont know is if this guy at Mushkin is all wet or if he is an engineer who knows what he is doing in that forum

By the 70mv he posted for DDR3 1800 that would suggest a FSB termination voltage of 2.18v!!!!!!!!!!!

AND

45nm CPU's are NEGATIVE value mv for GTL when mv is used instead of the multiplier

the way this guy has this set.. hes saying

1.1v (FSB TERM) x .635 GTL = .6985mv (rounded to .70mv)

The problem is, him using the multipler does not make sense for that input because

[vgbaron]

amy help at the RE forum?


http://vip.asus.com/forum/topic.aspx?bo ... uage=en-us

[vgbaron]

BTW, if you'd rather call it a night and sleep on it - fine w/me. I don't need FSX until Saturday.

If so, I'll be home tomorrow around 3PM California time and can check in then.

It goes w/o saying how much I appeciate the help, especially when I know you have other stuff to do.

You make the call -

[NickN]

Well... in the Asus forum they point back to the basic math I already know which is what is causing this confusion to begin with

http://www.xtremesystems.org/forums/sho ... p?t=202292

those formulas do not match what Muskin is doing with that setting

[NickN]

Vic.. lets try one set of values and if they do not work... go back to the 1600 settings I gave you and you confirmed stable @ 1T and PL6 till I get this sorted out...


calculating here...

[NickN]

OK Vic

Vcore 1.28, NB 1.52..

[vgbaron]

booted - running OCCT

[NickN]

OK....

If it doesn't pass and raising vcore and NB does not help,. set vcore back to 1.28v and NB to 1.52

set -50 (minus 50) for each GTL and retry

Short of that I need to contact Asus and find out WTF!! they did and why Mushkin is using a strait .635 multiplier instead of the right Intel engineering formula for GTL

[NickN]

You know.. its also possible those sticks just wont REALLY do 1800 under all applications. Mushkin just shows that for clockers and success in stability is based on simple benchmark stress tests for sales... that does not mean the sticks will run 1800 in all applications. FSX is a memory subsystem killer so you may be stuck at 1600 and the numbers we came up with earlier on those sticks

[NickN]

actually... try -45 first.. then -50

if you notice its MORE unstable as you increase then try setting FSB TERM to 1.10 and drop the 4 GTL settings to -35

Thats it.. short of communicating with Asus and Mushkin before trying again thats the best I come up with today

EDIT: Scatch that 1.1v test.. I cant believe FSB Term would want to be 1.10 in a 1800 clock, thats another thing that does not make sense about the Mushkin values. If anything that would want to go higher than 1.25

[vgbaron]

passed 30min OCCT cpu&ram test

-40 settings

trying 1N

[vgbaron]

Hmmm - when I tried to shut down to set the BIOS - got a bsod - which setting should I change?