Gigabyte GA-MA790FXT-UD5P Thread

[Barr3l Rid3r]

Fix some issues according to:
http://forums.tweaktown.com/f69/giga...tml#post359014
Links above uploaded to fileden.

You are the one, we just love you man!

I will give it a try right now.

It seems to fix a lot of the problems I was having, but I dont understand german very well.

[stasio]

You are the one, we just love you man!

I will give it a try right now.

It seems to fix a lot of the problems I was having, but I dont understand german very well.

Translate with :
http://www.microsofttranslator.com/Default.aspx
Edit:
First results from oc.net

Bugs:
-------------------------

CPU Clock Ratio:
can be set to x16.5(Core Boost Performance multi) even when the multi on the 1055T is locked at a maximum of 14.
when set to 16.5 it will be ignored and silently set to 14 without reseting the value in bios. maximum should be x14.

NB multiplier:
randomly changes to x10 when manually set to a x7 or x8.
x6 and x9 work fine.

ACC:
only 4 cores can be adjusted /are displayed

Core Performance Boost (CPB):
- when disabled the default Vcore is set to 1.350v
- when enabled the default Vcore is set to 1.475v

this can lead to some confusion when the Vcore is changed manually and the Core performance boost is then disabled or enabled.
maybe this can be seperated as well like the CPU multi?
example:


RAM:
Refresh Cycle Time: when set manually the value is set to a higher one, 90ns->110ns; 110ns->300ns; 160ns->300ns;

RAM:
here starts the real fun when trying to set anything manually....
my ram is rated 1600mz@1.65v@7-7-7-24-2T, it passes memtest+ 4.0 without errors,
but the OS crashes almost immediately and i had to reduce the frequency to 800mhz
and raised the Vdimm to 1.7v and the CR to 1T to get it anywhere near stable.
loser timings will make the ram even more unstable so it seams it behaves exactly like its not supposed to...
also i did some test to get higher freq stable but i would need insanely high voltage on the ram (like 1.85-1.9v)
to reach 1333mhz and even than it produces erros when tested in the OS (windows7 64bit, xubuntu 10.04 64bit)....


-------------------------
Fixed/Improved:
-------------------------

CPU core control:
all cores available

Cool&Quiet / Core Performance Boost(CPB):
- C&Q "enabled" / CPB "disabled" --> fixed
- C&Q "disabled" / CPB "disabled" --> fixed
- C&Q "disabled" / CPB "enabled" --> fixed

Vcore:
seems to be very stable now when CPB is enabled, used to jump araound all the time before.

RAM:
exact ram frequency is now shown when booting and not the value thats calculated from FSBxRam multi.

RAM multiplier:
randomly increases to x5.33 when manually set to x4. this seems to be fixed as well,
i could not reproduce it, but it might occure again since it was very random and hard to reproduce.

http://www.overclock.net/9869541-post42.html

[Barr3l Rid3r]

Translate with :
http://www.microsofttranslator.com/Default.aspx
Edit:
First results from oc.net

http://www.overclock.net/9869541-post42.html

Ok, but now I've seen that he's talking about the problems in F8H..

There is nothing about the F8j or F8k in that page...


tested the F8k and the stuck w/ CnQ always on is gone, but still impossible to use higher then 14x multi and elpida OC is bad in 6.66x memory divider (1600mhz divider I think)