OK...just built an i7 system, now what's going on? Only 4GB out of 6GB showing......

Just finished an i7 build.....some new parts, some old.

New parts:

i7 920
MSI X58M motherboard
6GB (3x2GB) Patriot Viper DDR3-1333 RAM



Old Parts:

4870 1GB video
2 x 1TB WD hard drives



Hooked everything up, fired it up, it lives but only showing 4GB of my 6GB of RAM in BIOS and in Windows (Vista 64.)


So, remove a stick from slot 3, now shows 2GB. Hmmmmm........

Put it back in, remove stick from slot 2, restart it and still shows 2GB. Hmmmmm.......


Put that stick back in and remove stick from slot one, which supposedly will prevent it from even starting, according to the manual anyway--and it fires up and shows 4GB of memory. Weird.

So, pull the stick from slot 2 and put the stick I removed from slot 1 into that slot, fire it back up and it shows, again, 4GB of memory.

OK.....so I remove ALL the memory, and put one stick into slot 1....the "required to be filled before all other slots" slot.

Computer starts for a few seconds, dies, restarts, and loops like that, all while beeping cannot locate RAM fault beep code.

So I switch the RAM around.....put every stick into slot 1....and it faults like it has no RAM.

Put any stick into slot 2 or 3 and it fires up. All sticks work in slots 2 and 3, no matter what's in slot 1.....


Sound like a bad motherboard? All the memory sticks work as long as they're not in the first slot.


And I'm using it right now as I post here....slot one empty, slots 2 and 3 populated, running in dual channel mode.

The majority of such problems arise from bent pins. It is happening much more often than before but no one seems to know why.

Have you lapped your CPU?

Quote:

---

Originally Posted by 003

Have you lapped your CPU?

---

No, see no need for that.

Don't worry, he's got his own theory which I think you just shot to pieces XD

Quote:

---

Originally Posted by MomijiTMO

Don't worry, he's got his own theory which I think you just shot to pieces XD

---

Correct. Which is not necessarily a good thing, this is a very frustrating problem that more and more people are experiencing on a wide range of motherboards, myself included.

try to change modules in their slots

So, if I understand it correctly from reading the post above, memory slot 1 is dead. You can either do an RMA or return it to the place of purchase. Possibly a better idea to just return it and get a new one to replace it.

Try a reseat of the CPU, add corect voltages for the rams, if no go RMA the board

Before you reseat, load up memtest+ and see if all 6gb are seen.
I had the same problem with my EVGA x58, however it turned out the problem was directly related to me using my Dell Perc5/i with the system once that was removed the system was able to fully utilize 6gb of ram.

Humminn55,
It's a pretty common issue (maybe too common) with i7, you can read my explanation to the problem here:
http://www.xtremesystems.org/forums/...95&postcount=5

I suggest you go with Leeghoofd's suggestions.

Good luck.

i think i have been read same problem from another guy within one month ago

Well, I've reseated the cpu twice now....memory volts are what they're supposed to be, at least according to the BIOS. Changed the sticks around endlessly.....still got a dead slot. Hate making the trip to exchange it, but guess I must.

Hate it even more tearing it down......that's actually the worst part.

Memory gap feature?

When i enabled this at 2gb in bios it showed as 4gb usuable in windows.

EDIT: that wouldn't explain the first slot being the only one that doesn't work, ignore me!

i had exactly the same problem with foxconn bloodrage, slot closest to cpu didn't work.
but.
i knew about weak pins and i was really extraordinary carefull installing cpu. btw i installed cpu only 2 times. and my pins was in ideal health.
so i returned board and got new one, gonna check soon if all slots will work nice. i hope this was not due to cpu uncore damage.

also i have my own theory about it :)
i have noticed that on first board latch on socket was very hard to lock, i need to apply a force to get it closed.
on second board latch closed very easy, without force. so this may be a real reason.

I also experience this exact same issue on my i7 rig... when I first built it, all 6GB was being recognized no problem, but one day I fired it up and it would only see 4GB, I then tried swapping ram around in different order, but nothing worked..

It is also the first slot on my mobo that fails to see the installed ram, so I like you guys had thought that it was a dead DIMM slot on the mobo, and went ahead and RMA'd the mobo, well they tested my board with 6GB and 12GB and it worked fine, so I got it back again, only to find that it was still only detecting 4GB..

I have two different sets of tri channel DDR3, and it does this with both sets.. I have also tested each module individually with memtest boot disc, and they all ran for 20+ passes without a single error at their rated spec...

I was still positive that I had a dead slot on my mobo, and that it was an intermittent fault, because some days I could turn on my system and in would see all of the installed RAM, but that would only last for a couple of hours at the most before going back to seeing 4GB, so I went and bought yet another UD5 mobo, and again the EXACT same problem happened on my brand new board!!!

I have tested my gear with as many different hardware configurations as possible, as I have 3 gfx cards, 2 sets of RAM, 2 UD5 mobos, 3 SATAII HDD's, 2 PSU's, and I also changed over all my SATA cables incase it was dodgy leads that was causing these problems, and no matter what hardware configuration I used this problem is always present, and the only piece of hardware that hasn't changed with each configuration is my processor...

I found a thread on the Corsair support forums, where 2 other people were having this problem with their rig, they were both using different mobos as well, an Asus RIIE, and P6T Deluxe I think it was, and one of the guys ended up changing his CPU and that sorted it out for him!!

I have been suspecting the IMC on my processor as the culprit as well, because like I said, I have changed over all other pieces of hardware, and the problems still persist...

I'm getting another i920 D0 next week hopefully, so then I will know for sure if it is my processor that is bricked.. though I am quite confident that it is...

Another thing that can cause this to happen is bent CPU socket pins.. I read a thread on OcUK forums where someone was experiencing this, and he had a bent pin in his socket, he straightened it out, and the problem dissapeared...

I wish that it was a bent pin that was causing my issues, because I hate RMA'ing hardware, but I have checked the sockets on both of my mobos 1000's of times, and there are no pins out of line at all on my boards...

Anyway, I will post back here to let you guys know how I get on with this issue...

Have any of you found a fix?

::edit::

@rioja, I to thought that it could have had something to do with the load plate in the CPU socket, because on both of my UD5's, they take quite a bit of pressure to close.. I was quite scared the first time I had to push down the bar to lock it in place... How could I go about loosening the load plate so it doesn't take the same amount of pressure to close?

Blunt you need to reseat the CPU a few times... that could solve it, there's prolly nothing wrong with the CPU... not the boards... nor the ram... just a bad contact

Hey man,

I have tried reseating the processor loads of times.. I don't know what else I can try..

It is weird because as I said, when I first built this rig, it would see all 6GB without any issues, then one day it started only seeing 4GB...

I don't think it has anything to do with reseating the processor, because it does this on both mobos, and as I said, I have tried to reseat the CPU loads of times and it doesn't make a difference...

I do notice something else that is strange in the BIOS though, the Round Trip Latency setting for my RAM is 62 on the first stick (the channel that isn't working) and then it is 58 for the other two sticks, someone on here told me that RTL should be first stick = x, second stick = x+2, and last stick should be x+4, anyway, I tried to manually set these, but still it only boots with 4GB...

Also, another reason why I don't believe it has anything to do with the processor needing reseated, is that certain days it does POST with 6GB, this maybe happens once a week at the most, but as soon as I do anything that puts a load on the system, it will crash, and then when it reboots it only sees 4GB again...

I want to try and loosen off the load plate to see if less pressure will sort this out, because it takes quite a bit of force to lock the load plate...

oh, i forgot to inform, i changed those bad mobo to new one, and all works fine
so it's no the issue with cpu, just the board
it had ideally looking pins on the socket, and reseating cpu didn't help
only rma is the way

and on new board socket latch was need same much force to close, so it is not the reason also

Well, I am 100% certain it isn't my mobo, for the simple fact that I have already RMA'd my first UD5, and also, my 2nd board does the EXACT same thing, and it is the same memory slot...

I guess it could be several things that cause this to happen, though I personally think that it is my CPU's IMC that is bad.. I think the problem I have is the same thing that the Fox said in the link he posted above...

I will RMA my processor on Tuesday anyway, and see what the outcome is of that...

I'll post back in this thread as to how I get on..

Oh, and I have now tried to reseat my CPU/HSF 4 times this evening, and nothing makes a blind bit of difference...

Quote:

---

Originally Posted by BLunTSmoKeR

Well, I am 100% certain it isn't my mobo, for the simple fact that I have already RMA'd my first UD5, and also, my 2nd board does the EXACT same thing, and it is the same memory slot...

I guess it could be several things that cause this to happen, though I personally think that it is my CPU's IMC that is bad.. I think the problem I have is the same thing that the Fox said in the link he posted above...

I will RMA my processor on Tuesday anyway, and see what the outcome is of that...

I'll post back in this thread as to how I get on..

Oh, and I have now tried to reseat my CPU/HSF 4 times this evening, and nothing makes a blind bit of difference...

---

A bad overclock can also cause it; if you've overclocked, back down to stock. Usually, when I've experienced this, it usually boils down to either too much vdimm or NB.

Quote:

---

Originally Posted by Zucker2k

A bad overclock can also cause it; if you've overclocked, back down to stock. Usually, when I've experienced this, it usually boils down to either too much vdimm or NB.

---

Thanks for the suggestion mate, but unfortunately this isn't the problem.. my system does this no matter how the BIOS is set..

I have even tried underclocking the processor, and having the RAM running at 1066, and 800MHz... nothing will make my system POST with 6GB of RAM...

The IMC on my processor is clearly damaged, I have tried absolutely everything that people suggest on various forums for this issue to no avail...

I understand that it can be a variety of things that can cause this problem to happen, like bent pins in the CPU socket, processor not being mounted properly, HSF being screwed down to tight, bad o/c settings, but it is none of these that is the cause of my problem..

I will be shipping my processor off tommorow or Wednesday, which will prove that it is my CPU that is at fault, and none of the above!!!

Well, I RMA'd my processor again, and there weren't any faults found, but the good guys at OcUK have sent me out a brand new boxed retail chip as a replacement anyway, even though my last chip was OEM!!

I got a PM from someone on here today, and they were experiencing the exact same problem as me, and they to replaced their mobo, RAM, and even their processor and the problems were still present, but someone suggested to him that it could be his HSF being to tight that was causing this problem, so he refitted his cooler and didn't have it so tight, and now his rig is seeing all 6GB and has been running perfectly for the last month...

Maybe it has something to do with my Zalman LGA1366 mounting bracket, so when this new processor arrives (it is supposed to be here today) I will install my TRUE now that I got the 1366 bolt thru kit for it, and hopefully that clears up the problem for me...

Quote:

---

Originally Posted by BLunTSmoKeR

Well, I RMA'd my processor again, and there weren't any faults found, but the good guys at OcUK have sent me out a brand new boxed retail chip as a replacement anyway, even though my last chip was OEM!!

I got a PM from someone on here today, and they were experiencing the exact same problem as me, and they to replaced their mobo, RAM, and even their processor and the problems were still present, but someone suggested to him that it could be his HSF being to tight that was causing this problem, so he refitted his cooler and didn't have it so tight, and now his rig is seeing all 6GB and has been running perfectly for the last month...

Maybe it has something to do with my Zalman LGA1366 mounting bracket, so when this new processor arrives (it is supposed to be here today) I will install my TRUE now that I got the 1366 bolt thru kit for it, and hopefully that clears up the problem for me...

---

Check your mounting standoffs.

It took me 3 months and a bounced RMA to realize that I had an extra stand-off that was shorting out a couple of ram pins from underneath. This effectively disabled 8 out of 16 ram slots on my workstation until I removed that standoff.

I have 9 stand offs on my mobo tray, and they are all being used...

Happens to me when i overclock, normally seems to mean that one of my channels doesn't like the timings i have set, even had it drop down to 2GB showing once.

@ BLunTSmoKeR

Have you had a chance at trying out the new solution for the RAM problem? My buddy has a similar problem so i wonder if the HSF(and the pressure) is the culprit.

Does the ram work in another computer?

Quote:

---

Originally Posted by Roliath

Before you reseat, load up memtest+ and see if all 6gb are seen.
I had the same problem with my EVGA x58, however it turned out the problem was directly related to me using my Dell Perc5/i with the system once that was removed the system was able to fully utilize 6gb of ram.

---

That's related to an SMBus problem. If you tape PCIe pins 5 & 6 on that PERC5 that problem will disappear.

I had the same problem, but only when trying to have my ram run at tCL 6. I first thought it was one of my sticks which was unable to run at this low latency. But today when I swapped out the heatsink (Did not re set the cpu, only installed a new sink). Then suddenly the problem were gone and all my dimms would work under C6.
Edit: Might add the first sink had the standard clip on seating. While my new one has a proper backplate and seating. So I'm presuming the new added pressure is what fixed my problem. The degrees is only few less, so highly doubt my problem was caused by heat.