Ram issues

[karambiatos]

I replaced my CPU cooler today, and afterwards, my PC would no longer start up, it started going into a boot loop, where it turns on stays on for 5 seconds and then turns off, and continues to do so, it does not post.

 

i figured out that having a ram stick in the 1st or 2nd slot causes this, the interesting thing is, it was not a problem until i replaced my CPU cooler.

 

i also tried resetting the bios, but it didnt work, after that i tried updating them to the newest version, im thinking its some kind of compatibility issue with the ram, since i had problems in the past where it would under clock it self to 1333mhz, or would cause a 2x boot loop, and then it would boot normally.

GPU: GTX570 x2

Mobo: Gigabyte Z68X-UD5-B3

CPU: Intel Core i7-2600k

RAM: 8GB DDR3- Corsair Vengeance 1600mhz (CMZ8GX3M2A1600C9)

 

PSU: 700W Cooler Master Silent Pro Gold

OS:Windows 7 Ultimate 64-bit

[Rhinox]

I had the "same" memory-kit (just CL-eight, stupid emoticons!)8), and I remember it did not play well with my mobo until I set all memory timings manually. Either SPD was somehow screwed, or bios did not detect it correctly. I can only recommend you to do the same: check corsair's web for specs and set it up manually (I think yours is 9-9-9-24 @1600).

 

Next, in some mobos you have to populate ram-slots in prescribed order. For example, in my case two ram-modules must be in 1st and 3rd dimm-slot. If I put them in 1st and 2nd, system does not boot.

Edited September 30, 2013 by Rhinox
Because stupid forum translates "8" + ")" as emoticon!

[EtherealN]

"Underclocking" to 1333mhz is not an incompatibility; you have a 68-series motherboard chipset. Intel specification for those is that everything above 1333mhz is overclocked, therefore it will never set memory to that speed on it's own no matter what the JEDEC profiles might say.

 

Did you clean the CPU surface of old thermal interface paste and reapply new after this cleaning had had time to properly dissipate? If there's old thermal interface paste still there, or a trapped layer of cleaning fluid on the heatspreader, what you are seeing might very well be a thermal stop - the UEFI is halting the computer in order to stop your CPU from (literally) burning.

 

Does your mobo have lights or number set viewvable so that you can see during which step it halts?

[Lange_666]

im thinking its some kind of compatibility issue with the ram, since i had problems in the past where it would under clock it self to 1333mhz

 

When i first made my current computer (look signature), i had a Gigabyte X58 UD3R motherboard which didn't run stable if i overclocked the standard 1333Mhz for the ram to 1600Mhz which i had fitted. It rebooted or entered a boot loop and then reset itself to 1333Mhz after that (with corsair 1600 dominator ram).

Finally i took it back to shop and asked for a replacement which they didn't want to do because like EtherealN says, anything higher then 1333Mhz was overclocked and because there it didn't show any problem. But after a long discussion with the board still hooked up on their test machine, it suddenly went into the boot loop (after rebooting it a dozen times with no luck to produce the same thing i had encountered at home) they agreed to change the board to my current Asus if i payed up for the difference in price and i did not have problem with it since (same CPU, same memory clocked at 1600).

 

I'm looking to replace my current system for something new and i'm still a bit wary about Gigabyte motherboards. For the moment i'm looking at either Asus again or MSI.

[Rhinox]

"Underclocking" to 1333mhz is not an incompatibility; you have a 68-series motherboard chipset. Intel specification for those is that everything above 1333mhz is overclocked, therefore it will never set memory to that speed on it's own no matter what the JEDEC profiles might say.

 

Not quite so. Z68-chipset has nothing to do with memory speed, because memory banks are directly connected to CPU. Therefore you will not find anything about memory speed in chipset specification.

 

On the other side, i5-specification allows ddr1333/ddr1600, and list of compatible memory modules goes even up to ddr2933.

[karambiatos]

im sorry guys, but did anyone actually read this part?

i figured out that having a ram stick in the 1st or 2nd slot causes this,

if i want to run the ram in dual channel it needs to be in ram slots 1 or 3, or 2 or 4, but whenever i put anything in ram slot 1 or 2, even if its only one ram stick in the computer, it keeps boot looping.

 

Did you clean the CPU surface of old thermal interface paste and reapply new after this cleaning had had time to properly dissipate? If there's old thermal interface paste still there, or a trapped layer of cleaning fluid on the heatspreader, what you are seeing might very well be a thermal stop - the UEFI is halting the computer in order to stop your CPU from (literally) burning.

I dont have uefi, and if the issue was a thermal stop, how come it boots up fine if i put my ram in ports 3 and 4, but not any combination involving ports 1 and 2.

 

Does your mobo have lights or number set viewvable so that you can see during which step it halts?

it halts right before posting.

 

 

Also if anything about 1333mhz is over clocked why would manufactureres list ram that is higher than that as compatible with the motherboard?

http://www.corsair.com/us/learn_n_explore/

Edited September 30, 2013 by karambiatos

[EtherealN]

Not quite so. Z68-chipset has nothing to do with memory speed, because memory banks are directly connected to CPU. Therefore you will not find anything about memory speed in chipset specification.

 

On the other side, i5-specification allows ddr1333/ddr1600, and list of compatible memory modules goes even up to ddr2933.

 

Yeah, sorry, I misspoke, yes; the UEFIs/BIOS interrogates the controller (in this case CPU) for compatible settings, and configures according to this. On a Z68 release this means "2nd gen Core" and, in this specific case, i7-2600k.

 

And: http://ark.intel.com/products/52214

 

It is correct that if using a "3rd gen" (Ivy Bridge) CPU, this level becomes 1600.

[EtherealN]

Also if anything about 1333mhz is over clocked why would manufactureres list ram that is higher than that as compatible with the motherboard?

http://www.corsair.com/us/learn_n_explore/

 

If you look closely at the memory compatibility charts (assuming your vendor supplied a proper one), you'll note supports above 1333 and/or 1600 (depending) as "OC", for "overclock". What it means is that they have tested these memory sticks at their design specifications and they work.

 

But the memory controller and BIOS/UEFI will still consider this an overclocked memory. (I have the same processor as you do, and I similarly had to reconfigure my 1600MHz memories from 1333 - but my manual warned me about this. It is normal.)

 

Have you run Memterst86+? I would recommend doing so with the sticks in the operational positoin first, just to make sure there's nothing wrong with them: http://www.memtest.org/

 

It might be total coincidence, I've had mysterious things like that happen to me a few times on previous machines; sometimes the memory controller (depending on it's status as well as that of mobo connections etcetera) is more sensitive to faults on some ports; similarly support for given configurations of sticks can sometimes only be for specific port setups. (Black magic.) Since it worked previously, it shouldn't be a support issue, but it is possible (if remote) that something happened on first boot after the reseat of the cooler that damaged the controller and made it more sensitive.

 

Either way, a full memtest run should probably be your first step.

[SkateZilla]

I know with AMD, when using all 4 slots it requires the speed to be lowered to maintain stability with the IMC; experienced AMD OCers like myself know how to up the IMC Voltages to Allow Max speed on all 4 slots.

 

However, not sure about Intel Chipsets.

 

Also, Try removing all but 1 stick in slot 1. But if it's Cutting the Power when a stick is in slot 1 it may be a BIOS self test fail

 

If you cant boot with 1 stick in slot 1, try a stick that didnt give your problems before.

 

if it still wont boot, then the Mainboard has a bad slot, should prolly look into Warranty RMA.

[karambiatos]

so i did 10pases with memtest, and no errors were detected.

 

ill try cleaning the ports more thoroughly.

Edited October 2, 2013 by karambiatos