problem with gigabyte GA-X58A-UD3R

I have had this a couple of times now.

Setting the "performance enhance" setting to standard seems to have fixed it.

On default non overclocked settings the pc would sometimes blue screen and also loose 2GB if extreme setting was used.

it didn't make a difference what setting i use
 
Great... had 3 crashes and blue screens last night.

Now its just done it again and only 4gb of ram ( all on default settings ).

When i do a mem check all tests are fine with the full 6gb ( when it detects 6gb ).


What the hell am i supposed to do now? taking the gear back will result in a "no fault found" i suspect.

EDIT: After some more searching there is a new BIOS ( F7 ) update that appears to deal with the memory issues. Ive flashed and the layout for all the memory and overclocking settings have changed. I shall report back on wether i get any more issues when i have used the machine for a few days. The new BIOS is dated 2nd of august 2010 so heres hoping this will sort it.
 
Last edited:
For anyone looking for BIOS:

http://forums.tweaktown.com/gigabyte/28441-gigabyte-latest-bios.html

The first post is kept updated.

THESE ARE BETA BIOS!!!!!
THESE ARE BETA BIOS!!!!!
THESE ARE BETA BIOS!!!!!
THESE ARE BETA BIOS!!!!!

I will be flashing my X58A-UD3R Rev 2.0
I will keep you updated.

EDIT: Have updated mine fine with the beta BIOS. No issues yet.
6 Gb on 4 reboots.
 
Last edited:
Well ive had 6gb each time BUT i fail prime 95 after about 5 mins now AND the pc crashed out with a IRQ_EQUAL_OR_NOT_LESS_THAN error.

I did a memtest from a boot cd and that failed with memory errors after about an hour and a half.

I have just tested each one of the 3 mem sticks on its own and suprise suprise... they all passed :mad:

So what the hell do i try now?


EDIT: It seems to be stable in dual channel mode with dimms in slots 1 and 3 at maxx speeds. Well its done an hour of prime 95 anyway.

Im gonna stick the 3rd back in and see what happens.
 
Last edited:
Great... 4 mins into prime and it fails again.

At this rate it HAS to be the board.

Ive tested each stick one by one in each slot and im running everything at default settings and its not stable.
 
Ok update time.

I think i have it sorted now. Schoolboy error it seems.

I was unaware that to use 1600 memory speeds you had to up the QPI/VTT voltage a little. :p

I have upped to 1.275 and i am now stable in triple channel mode at stock speeds at 800mhz memory clocks.

I just assumed the auto detect SPD would set the voltages required. Seems it only sets the DRAM voltage itself and my memory controller was hungry for juice when stressed hence the failures.
 
Last edited:
It seems to be a fault with the foxconn sockets and aftermarket coolers. I have been having the very same issue on my P6X58D-E, i am using a Venomous X cooler and was having bsods and only showing up as 4gb in bios. I have now gone back to the standard cooler and its fine, no more bsods and its showing the full amount of ram. It is a little strange as i wasnt getting the problem on my ud5...
I have also read applying thermal insulation tape to any metal heatinks on the back works, by stopping metal contact on the rear.
 
i really need help with what to change to, i'm replacing it with a different make but don't know what mb doesn't suffer with the same problem
 
Well, tried the new BIOS, kept getting BSOD Memory_management that went away when I reverted the BIOS back to the most recent one on the website.
 
Already have. no issues.
except what is the point of having a motherboard designed for overclocking when you only have the stock cooler?
 
I had this problem on my UD5 and H50 when I first got it, turned out I had a bit of dust in the socket, god knows how it got there, but i blasted it out, reseated the CPU and cooler and had no problems. Not sure this adds anything but maybe it will help some people, recheck the socket and reseat the CPU, might help.

Hawker
 
Tried it and didnt make a difference.

I inspected the pins and marks on the cpu and all is as it should be.

This has got to be a board/bios fault and im getting very annoyed now.

The thing is it seems to be intermittant, so upping a voltage here and there seems to fix it then it crops up again a few days later.

So taking the stuff back and letting it be tested is going to be a recipe for disaster, no fault found sir! :mad:
 
please don't say that I've just been given a return number for them to test the cpu, ram and mb, if they can't find a fault I've got to pay deliver cost of £15 + vat
 
Back
Top Bottom