GA-X58A-UD7 Not Recognizing All RAM

[ccage90]

Hello all,

I have had trouble getting the MB to recognize all 3 DDR3 RAM sticks in the correct slots. Under BIOS MIT Current Status will show 4Gb in DIMM slots 1, 3, 5. But on the Enabled Slot line it shows 4Gb in slots 1 and 4 for a total of 8 Gb. Under windows, it will show 12 Gb (8 Gb usable).

I have read about this issue and that it can be due to bent CPU socket pins. I have removed the CPU and didn't find bent pins. But I wouldn't say I have a trained eye for this.

I have read that this can be due to BIOS settings that need to be tweaked to tune the voltage and timing settings. I haven't tried much of that tuning yet.

I have tried to get help from Gigabyte support. They have given me suggested steps for troubleshooting and I will detail those results below. Any words of wisdom on diagnosis or next steps I should try are appreciated.

System Specs:
- GIGABYTE GA-X58A-UD7 LGA 1366 Intel X58 SATA 6Gb/s USB 3.0 ATX Intel Motherboard
- OCZ Reaper Edition 12GB (3 x 4GB) 240-Pin DDR3 SDRAM DDR3 1333 (PC3 10666) OCZ3RPR1333C9LV12GK
- Intel Core i7-970 Gulftown 3.2GHz LGA 1366 130W Six-Core Desktop Processor

Motherboard UD7 v2.0 - BIOS ver. FB

Currently running BIOS optimized defaults with only setting the Memory voltage to 1.64.

OCZ Memory ratings for these chips:
1.65v, 9 9 9 20

I will attach a picture that shows the MIT status values and illustrate the problem. Sorry that it is displaying upside down after upload.

It is showing that the memory is a little higher than 1333 and timing is 9,9,9,21 instead of 20.

Thanks for any assistance.

.chris

[ccage90]

Here are the testing steps that Gigabyte support asked me to try and my notes on results. I think they were trying to determine if one slot is bad. Too much detail probably.

1) Restore default settings and set voltage only
- Load Fail Safe Mode
- Load Optimized Defaults
- Set Voltage to 1.64
- Save settings and restart
- Results: Still shows 8Gb for 3 chips under MIT status (picture taken)

2) Test with 2 memory sticks
- Slots 1 + 3
- Sticks 1 + 2
- (remove stick 3 from slot 5)
- Results:
- 1st boot: 8 Gb Normal
- 2nd boot: 4 Gb
- 3rd boot: "Recovering Lost DRAM" -> Auto Reboot
- 4th boot: Windows -> 8Gb (4Gb usable)

3) Test with 2 memory sticks
- Slots 1 + 3
- Sticks 1 + 3
- (Swapped stick 3 into slot 3)
- Results:
- 1st boot: 4 Gb
- 2nd: 4 Gb -> Recovering lost DRAM -> Auto Reboot
- 3rd: 4 Gb -> Recovering lost DRAM -> Auto Reboot
- 4th: 8 Gb -> Windows -> Shows 8 Gb!
- 5th: Doesn't show mem size -> straight to Windows -> Shows 8Gb again
- 6th: Into Bios -> MIT status whos 8 Gb Normal
- 7th: Windows -> 8 Gb Normal still

4) Test with 3 slots again
- Put stick 2 into Slot 5
- Results:
- 1st: 12 Gb -> Into BIOS -> MIT shows 12Gb and 3 slots Normal
- 2nd: 12 Gb -> Windows -> 12 Gb Normal
- 3rd: Seemed to be starting window -> then beep - auto reboot!?!
- 4th: 12 Gb -> Windows -> 12 Gb Normal
- 5th: 12 Gb -> "Recovering Lost DRAM" -> Auto Reboot
- 6th: 8 Gb -> Recovering -> Auto Reboot
- 7th: 8 Gb -> Recovering -> Auto Reboot
- 8th: 8 Gb -> Entered BIOS -> Took picture of MIT status showing 3
sticks 2 recognized
- 9th: 12 Gb -> Windows -> 12 Gb Normal
- 10th: Windows -> 12 Gb Normal
- 11th: Entered BIOS -> MIT shows 8Gb? - 3 sticks, 2 recognized
- 12th: 8 Gb -> Recovering Lost DRAM -> Auto Reboot
- 13th: 12 Gb -> Windows -> 12 Gb Normal
- 14th: Recovering Lost DRAM -> Auto Reboot
- 15th: 8 Gb -> Entered BIOS -> MIT shows 8Gb - 3 sticks, 2 recognized
- 16th: 8 Gb -> Recovering -> auto Reboot
- 17th: 12 Gb -> Windows -> 12 Gb Normal

Conclusion: still too unstable

[wazza300]

what qpi/vtt voltage are you using and are you at stock cpu clocks? try with 1.35000v for qpi/vtt and you may need more,set dram voltage to 1.66 for 1.65 in windows

[ccage90]

QPI/VTT currently at 1.175V Auto.

Yes, I'm at stock CPU. I haven't tried tweaking anything yet. Just want to get this stable with all RAM first.

I will try setting the QPI/VTT and DRAM voltage and send update.

Thanks for reply.

-chris

what qpi/vtt voltage are you using and are you at stock cpu clocks? try with 1.35000v for qpi/vtt and you may need more,set dram voltage to 1.66 for 1.65 in windows

[Acebmxer]

Here are the testing steps that Gigabyte support asked me to try and my notes on results. I think they were trying to determine if one slot is bad. Too much detail probably.

1) Restore default settings and set voltage only
- Load Fail Safe Mode
- Load Optimized Defaults
- Set Voltage to 1.64
- Save settings and restart
- Results: Still shows 8Gb for 3 chips under MIT status (picture taken)

The correct way is enter bios choose load optimized defaults press F10 save and exit.
Renter bios set your hdd's for raid or ahci if needed, Set voltages for ram, disable boot logo if you want. Press F10 save and exit.
Let computer boot into windows. Choose shut down or reboot renter bios and then make O.C changes Press F10 save and exit.

Also try removing cpu checking for bent or broken pins. if everything is ok remount your cpu cooler make sure to fully tighten it but do not over tighten it. (Memory controller is in the cpu)

[ccage90]

Tried setting QPI/VTT = 1.355 and 1.375, DRAM Voltage to 1.66

Same behavior: 8 of 12 Gb and showing slots 1 and 4 enabled instead of 1,3,5

Tried setting timing to 9,9,9,20 and no change.

Is it still possible that this can be fixed by BIOS settings or this is pointing to a problem with the CPU seating? Why would bad seated CPU not show other symptoms of trouble? I can run windows 7 OK for some time with single 4Gb RAM chip though I haven't stress-tested yet. Seems like bent pins or bad seating would cause random issues, not just a RAM issue...?

I've noticed memory frequency shown in the BIOS MIT status view is 1.349.88 Mhz (10X). Is this correct or should that be showing 1333?

Also, DRAM Voltage on BIOS status was still showing 1.648 after above tweaks. Is that far enough from 1.65 to cause this problem?

Thanks again.

[wazza300]

1.648 is the right amount in windows,have you tried with more qpi/vtt? try 1.4-1.44v and see if the full 12gb show up,

test each ram module one by one in each white slot one at a time,it will boot with one stick in any white ram slot,you cant test the blue slots as it wont boot from them unless 12gb are used,atleast you will know whether or not you have a dead/faulty ram stick or slot

[SolidBladez]

Try loosening your CPU heatsink just a bit. And like you said, it could be some bent pins. So if loosening the heatsink doesn't help, try checking for that.

[xdillweedx]

I'm sure that you are pretty confident in building systems, but I have to ask did you take out the ram and reseat it several times? Whenever, I'm building a new computer with a new motherboard and ram, I'll put the ram in, take it out and put it back in several times to make sure that all the ram pins are fully contacting the ram. Whenever a stick of ram is not recognized this will usually solve the problem. (at least is has for me)

All that said, it could quite possibly be something else :)

Good luck

[Psycho101]

That's a good suggestion. You can also remove the RAM sticks one by one and check for contact. Look at the gold contacts. They should have small scratch marks where they've rubbed the slot pins on insertion. If any of the contacts are mark free, this could indicate a contact problem.