6GB of DDR3 only showing up as 4GB... P6TSE/G.Skill in Tri

[MmmDieselFumes]

Ok, I just built this computer about two weeks ago. Every couple of days I'm having an issue where all of the sudden, one of my sticks of ram disappears from windows (though not from CPUID):

http://i302.photobucket.com/albums/nn114/purecountryrebel/Untitled-1-7.jpg

And according to CPUID, all three dimms are full:

http://i302.photobucket.com/albums/nn114/purecountryrebel/Untitled-2-1.jpg

To resolve this issue, I open up the case, remove all of the sticks of ram, and reinstall them. Then I turn on the computer, and windows and everything recognizes them just fine for a day or so. I have no idea whether this is a ram, or mobo issue. When windows see's all 3 sticks (totaling 6GBs), and I run HyperPI on 32m and Prime95, it stress' it just fine with no problems.

Then all of the sudden, out of nowhere, I turn on the computer, and windows only "see's" 4GBs, while CPUID see's all 6!

This is starting to drive me bat s*** insane... Does anyone have any idea what's going on here, or how to diagnose further?

[Iam99x]

That's strange. Have you tried to switch them around in the slots?

[MmmDieselFumes]

Yeah, I've tried flopping them around, and even only running one at a time... They all "work" until one of them doesn't, but I can't tell which one isn't working.

[Elfie]

Check out the post from GSKILL TECH http://www.gskill.us/forum/showthread.php?t=4059&page=2

[MmmDieselFumes]

Not really having a cold boot issue (seeing as how I can boot into windows, have no BSOD, or endless loops). But thanks for the link. Has some interesting info.

[dominick32]

You may have a single faulty stick. I know you already stated that you already moved "swapped" them around. But I have already seen your problem in 2 different cases, and each case was caused by having a faulty stick. Even though they are listed in CPUZ there still may be one stick of memory that is slightly finicky and return material. I would recommend testing each stick in DIMM A1 the first slot, one by one and find out if one stick does not allow the rig to boot up. Again, this is just to cover ALL of our bases first. Than run memtest on each stick individually for 3 minutes or so to get a base reading. If you did truly already test them, thats a different story.

-D

[MmmDieselFumes]

You read my mind Dom. Started doing that exact thing right before you posted. Had some interesting discoveries.

With 4gb registering in windows, and the full 6gb in cpuid, I restarted the computer and went into the bios. Only 4gb registered in the bios, so I figured I'd start testing individual sticks. So I opened up the case, and as my fingers were touching the very right hand stick (c1) I remembered I needed to check out the AHCI setting in the bios (thought I remembered seeing it in IDE Enhanced mode instead of AHCI). So anyways, my fingers just barely moved the last right hand stick, I put it back together to boot into the bios. Lo and behold, all 6gb are now registering. I load up windows, and here too, everything now starts up and shows the full 6gb as usual.

So I start to think I have a finicky stick, and proceed to check each individual stick. I tried all 3 sticks in the A1 slot, and boot up into windows. No problem, all 3 sticks come up as 2gb sticks in the BIOS and in windows. Strange I think to myself.

I then put all 3 sticks back in the correct spots (a1,b1,c1) and load into memtest86 from my usb. Ran each sticks for 15mins a pop with zero errors, so I assume they all check out...

Anyhow, I just loaded all 3 back in, bios and windows and all my programs are recognizing it as 6gb, so I really don't know where that leaves me. I think the stick that started working when I barely moved it is trying to tell me something, but whether it points to a faulty mobo or stick, I have no idea, because everything is running fine for now once again...

[Perseus]

I had the exact same problem last night when I tried to slap in some some Dominator GT. My system not reporting the installed amount of RAM properly was caused by trying to run XMP (7-7-7-20 1t 1600) without tweaking other settings to match. I don't pretend to understand all the ins and outs yet, but setting DIMM voltage to 1.65, QPI to 1.35 and bus speed to 1333 (in manual mode) solved the 4 out of 6gig reported problem.

[jamsde]

I'm having the EXACT same problem with Kingston Hyperx DDR3 2000 triple channel ram. I called ASUS and their tech told me the CPU had a bad RAM CONTROLLER, and that he hears of this problem somewhat often. Anyway, I RMA'd my CPU and newegg gave me a full refund.. new CPU is arriving tomorrow or Tuesday, i'll let you know if adding a new CPU fixes it or not.. Til then i'm swapping sticks again... gahhhh

[jamsde]

I think the ASUS tech support guy I talked to is nutz.. i tested each stick in slot A1, all test fine.. I test a stick in B1, ALL FAIL in memtest.. Same stick in C1, tests fine no errors.. test the other two sticks in B1, all failure...
Sounds like a bad slot to me? Thoughts? Could it still be the CPU Ram controller?

[jamsde]

Ok i'm 99.9% sure it's a motherboard problem, I put a stick in A1, A2, and C1... ran full round of mem test, no problems, booted to windows, Ran Prime95 for 5 mins, no problems... gotta be a bad ram slot on the board...

[baditude_df]

I just confirmed this on my RampageIIExtreme tonight. Reading 4 out of 6GB in all but CPUID. Started troubleshooting by trying the same stick in each of the DIMM slots. A1 booted fine, B1 no boot, C1 booted fine. Just for S&G I tried the other two sticks in B1 and no go. Pretty quick answer to my questions. Guess this board gets RMA'd. Oh well, system runs fine in Dual Channel till I find a temp board.

[Perseus]

I don't know if this will help you guys, but my problem turned out to be dead simple. All I needed to do was upgrade my BIOS to 0901 from 0504. Changes that would kick me back down to 4gig now work as they should, with no change to reported RAM.

[baditude_df]

Yeah, I could always try that. I spent all of last night clearing BIOS, changing clocks, and changing DIMM placement etc. Nothing worked. I'll give it a go with another BIOS.

[kchris]

Don't bother with Windows when it comes to recognizing hardware. If your BIOS and advanced softwares(such as CPUID) know its there, its there.

[baditude_df]

The only place that recognizes that all 3 DIMM slots are full of 2GB DIMMS is CPUZ. BIOS Post shows 4, task manager shows 4, Everest shows memory utilization between 68 and 75% while crunching Rosetta which is WAY too high. Is should be 50% or below.
System will simply not boot with a single DIMM in the B1 DIMM slot. A1 and C1 yes, but not B1.

[Perseus]

Don't bother with Windows when it comes to recognizing hardware.

:D You're joking, right?

[dominick32]

All I can say is: Judging by how many of you that have come across this problem it looks as though it is not simply just a coincidence. This looks to be a common occurence, I have no words of wisdom other than I feel bad for all of you guys. I know what a PITA it is to take apart an entire gaming rig to send in RMA's after waiting weeks and saving up for your hardware. Good luck and keep this thread updated with info.

-D

[Kryten451]

I just had the same problem with my X58 gigabyte / 6GB OCZ triple channel.. In my case I re-seated the memory (once again).. and it read 6 gig..

I cannot prove this but I think these mechanical slots for the DDR sticks are marginal because the memory slot Levers do not hold the sticks into the slot with a positive "lock", and the sticks "lift" back out from the pressure of the contact areas.. You could perhaps double and triple check, or put on a strap of some kind to compensate..

[Drew@PSU]

I have a R2E with something approaching this. I've only had it running for a week or two, but after I started messing with the RAM I noticed a stick seems to 'drop out' so to speak. I haven't had any boot issues, so I'm thinking it's just my inexpert ocing. I will keep an eye on this thread though. Best of luck to all involved.
-Drew

[baditude_df]

YEah, I'm running in Dual Channel right now. 4GB is enough, but definitely don't like the fact that the Mobo is not running the way it should.
No amount of tinkering can get my 6GB back. Time for RMA.

[MmmDieselFumes]

Sorry haven't posted in a while, been busy helping a friend drop his mercedes diesel in his toyota pickup... Talk about a headache!

Anyways, booted up my gaming rig after like 2 weeks of having not turned it on, and lo and behold, it only showed 4gb again.

So I opened the case, jiggled the c1 stick around, booted back into windows, and it still showed 4gb.

I then removed the c1 stick, looked into bios, and it then showed only 2gb of ram, so I think to myself, it has to be either a bad stick or socket at a1 or b1.

I took out b1, booted into bios and it showed 2gb, so I figure a1 is good, and b1 is my culprit. So I slap the stick back into b1, and boot up again, and it shows 4gb. I then start getting angry because the dog gone problem is still eluding me. So I finally put in the last stick into c1, and it shows up as the proper 6gb, and so I sit here now typing this, with everything "working properly" for now.

I figure it has to be an intermittent issue with the b1 socket. I'll have to run memtest with sticks in b1 socket and see what it shows me. Though I doubt that I'll see anything because the problem seems to go away for a while once you reseat the ram.

And the part that really blows is that just when I decided I'd go ahead and RMA it to newegg, I checked, and now it's been like 35 days since I ordered all these parts, so I guess I'm screwed blued and tattooed huh?

[MmmDieselFumes]

I don't know if this will help you guys, but my problem turned out to be dead simple. All I needed to do was upgrade my BIOS to 0901 from 0504. Changes that would kick me back down to 4gig now work as they should, with no change to reported RAM.

Yeah, I could always try that. I spent all of last night clearing BIOS, changing clocks, and changing DIMM placement etc. Nothing worked. I'll give it a go with another BIOS.

Yeah I'm gonna try updating from 0704 to 0805 right now and see where that get's me. Guess I'll just have to watch, wait and see...

[MmmDieselFumes]

And also, instead of starting a new thread, figured I'd just ask here, but anyways, are these temp ranges in the normal when running prime95 torture test on 8 threads with round-off checking, with no overclocking using a stock intel heatsink (and 5x120mm case fans)...?

http://i302.photobucket.com/albums/nn114/purecountryrebel/Capture.jpg

[dominick32]

And also, instead of starting a new thread, figured I'd just ask here, but anyways, are these temp ranges in the normal when running prime95 torture test on 8 threads with round-off checking, with no overclocking using a stock intel heatsink (and 5x120mm case fans)...?

http://i302.photobucket.com/albums/nn114/purecountryrebel/Capture.jpg

At 2.8 GHz, I would say possible? Yes. Realistic if you setup your board to generic specs? Hell no! I have a strange feeling your board is defaulting to high stock vcore.

Word of advice, I would go into the bios and manually select the option for vcore. Turn your vcore from auto to manual and set it at 1.08v. This will substantially lower your CPU and undervolt it a bit under VID usually. Your temps should drop to around 55 to 65 under load instead of the disturbing 85C. Something else to keep your temps even cooler at low vcore and low cpu speed is disabling HT (Hyperthreading). To do this, go into the bios by pressing DEL when you boot the computer and select the setting for CPU OPTIONS. Disable the setting for Intel HT Technology. On your next boot, this will 99% of the time give you on average a decrease of 10*C per core and increase your theoretical overclock by 100 to 200 mhz. The second option is only for people that dont need HT. The only real reason you would need HT on a Core i7-920 is if you have specifically tuned and HT optimized software. In many instances, games and applications will have a greater benefit with HT disabled because of added latency generated from the phsyical core instructions when a piece of software is not HT optimized. I honestly love overclocking with a passion! But I love people that choose to undervolt also so this is a great chance to experiment with how low she will go. lol ;)

Since you are choosing to run at such an extremely low CPU speed of 2.8 GHz and you are on the stock Intel HSF, this is an excellent opportunity for you to get into undervolting and/or finding the lowest stable vcore to run at that low frequency.

So, by starting with 1.08 vcore and disabling HT I can guarantee your temps will drop around 20*C right off the bat from both idle and load and the PC will run better than ever assuming your stock vcore selection is much higher than VID at default. :) Remember, this undervolting and HT disabling trick is only a temporary fix until you get yourself a Venomous X Heatsink with dual push/pull panaflo's. At that time I expect you to be at 4.1+ Ghz all day long. I can help you out either way dude. One word of advice, as long as you are running the Intel stock HSF do not OC with anywhere near those temps.

Cheers.

Dom

[nfriedman4]

New here - but was having exactly the same issues you were on a very similar build. I was pulling my hair out trying to figure out how to lower both the TEMP and figure out why windows was not recognizing all of my mem....ran all the tests, did all the reslotting, bios updates, etc.


First off - the memory...I was going INSANE trying to figure out my own problem with it. Alot of people suggest on boards across the internet that it may be a Mobo, slot, or mem issue. If you've checked every stick and they are all seated properly, then I do have a suggestion...reseat your CPU processor. Make sure that there is good contact between the CPU and the pins. I actually had a SMIDGE of thermal grease on a pin on the bottom of my CPU and it was causing the issues in reading the mem. I cleaned it with a bit of Arcticlean on a coffee filter, reseated it (making sure the CPU was perfectly in place) rebooted and on startup Windows recognized all 8 GB.


Now your Temps - I did so much research on it my eyes almost exploded. My temps would sit at 34c but then as soon as P95 started torch up to the 80's and I couldn't for the life of me figure out why. I eventually broke down and spent the $30 to junk the Intel Stock and go with the Cooler Master Hyper 212. No mods to the voltage or anything and the highest my CPU gets on full load (non OC'ed) is 45c after 8 thread FFT's test. Honestly, I totally agree with the post mentioning to lower your voltages, but why get a porche and treat it like a hyundia - the CM Hyper 212+ has made me forget all together about my temps and even opens the door for possible overclocking in the future. If you have the room in your case (fairly large) and don't mind taking out the Mobo one time (hey it gives you a chance to remanage your cables, etc) then I'd strongly consider.

[dominick32]

New here - but was having exactly the same issues you were on a very similar build. I was pulling my hair out trying to figure out how to lower both the TEMP and figure out why windows was not recognizing all of my mem....ran all the tests, did all the reslotting, bios updates, etc.


First off - the memory...I was going INSANE trying to figure out my own problem with it. Alot of people suggest on boards across the internet that it may be a Mobo, slot, or mem issue. If you've checked every stick and they are all seated properly, then I do have a suggestion...reseat your CPU processor. Make sure that there is good contact between the CPU and the pins. I actually had a SMIDGE of thermal grease on a pin on the bottom of my CPU and it was causing the issues in reading the mem. I cleaned it with a bit of Arcticlean on a coffee filter, reseated it (making sure the CPU was perfectly in place) rebooted and on startup Windows recognized all 8 GB.


Now your Temps - I did so much research on it my eyes almost exploded. My temps would sit at 34c but then as soon as P95 started torch up to the 80's and I couldn't for the life of me figure out why. I eventually broke down and spent the $30 to junk the Intel Stock and go with the Cooler Master Hyper 212. No mods to the voltage or anything and the highest my CPU gets on full load (non OC'ed) is 45c after 8 thread FFT's test. Honestly, I totally agree with the post mentioning to lower your voltages, but why get a porche and treat it like a hyundia - the CM Hyper 212+ has made me forget all together about my temps and even opens the door for possible overclocking in the future. If you have the room in your case (fairly large) and don't mind taking out the Mobo one time (hey it gives you a chance to remanage your cables, etc) then I'd strongly consider.

I agree with you as well. I was more stating that since he does not want to OC and he is at such a low clockspeed of 2.8 Ghz, he is throwing way too much juice into it. Not to mention, if he plans on staying that close to stock core speed he might as well undervolt and disable HT since he is on the stock cooler. If he was on high end air cooling like yourself, I would tell him to go 200 X 20, 1.30v, and reboot the rig at 4.0 Ghz all day long bc he would be essentially wasting a ton of available oc clockspeed. The i7-920 is golden, and one of the easiest if not the best 4.0 GHz capable quad cores on the market. :)

[MmmDieselFumes]

Jeeze, sorry my posts are so few and far in between. Been crazy busy lately, and no time for my hobbies lol.

Anyhow, thanks a TON for the help Dom. Thanks for the input too nfried.

But anyways, I just turned the rig on after letting it sit since thursday night just now, and surprise, surprise, it only showed 4gb (and that's AFTER having updated to that newest bios, 0805). So it's STILL having issues. On a whim, I went ahead and reseated the middle (b1) stick. That seemed to do the trick, and all 6gigs showed up.

And I went ahead and took your advice on changing the vcore and turning off HT. BOY HOWDY, WHAT A DIFFERENCE! Though on my little sidebar, it now only shows 4cores, which is ok since I assume HT just makes an addition 4 "virtual" cores. While running prime95 on 4cores, Core Temp shows my low idle @ 39-44c, and my load high temp for 15mins has been at a peak of 58c!

http://i302.photobucket.com/albums/nn114/purecountryrebel/Capture-1.jpg

So that solved my high temp issues splendidly! So now, I'm really just concerned about the whole original ram issue (for now. I'm sure other issues will pop into my head lol)... Any other thoughts on that front?

[dominick32]

Jeeze, sorry my posts are so few and far in between. Been crazy busy lately, and no time for my hobbies lol.

Anyhow, thanks a TON for the help Dom. Thanks for the input too nfried.

But anyways, I just turned the rig on after letting it sit since thursday night just now, and surprise, surprise, it only showed 4gb (and that's AFTER having updated to that newest bios, 0805). So it's STILL having issues. On a whim, I went ahead and reseated the middle (b1) stick. That seemed to do the trick, and all 6gigs showed up.

And I went ahead and took your advice on changing the vcore and turning off HT. BOY HOWDY, WHAT A DIFFERENCE! Though on my little sidebar, it now only shows 4cores, which is ok since I assume HT just makes an addition 4 "virtual" cores. While running prime95 on 4cores, Core Temp shows my low idle @ 39-44c, and my load high temp for 15mins has been at a peak of 58c!

http://i302.photobucket.com/albums/nn114/purecountryrebel/Capture-1.jpg

So that solved my high temp issues splendidly! So now, I'm really just concerned about the whole original ram issue (for now. I'm sure other issues will pop into my head lol)... Any other thoughts on that front?

Wow look at those temp differences. just as I predicted. I am glad that I got your loaded CPU temps down to 58*C from 84*C! LOL ;) Believe me, I have owned so many 920 C0's and D0's I couldve probably worked as an Intel Distributor.. I know the ins and outs of these buggers like mark mcguire knows steroids. :)

In any event. Now remember unless you need to undervolt for a word processing or office rig (strictly office type apps, multimedia, no gaming, no encoding, no heavy processing) this is only a temporary fix. As I stated previously, you need to remember you are running a golden Core i7-920 D0 processor, this is one of, if not the best 4.0 GHz oc quad core ever to exist. So, I recommend you buying a Venomous X high end CPU cooler with 2 fans in a push pull configuration. After setting up your new air cooling setup you should go back to selecting 1.30vcore and go 200 X 20 for 4.0 GHz 24/7 as a daily overclock.

I am a big fan of disabling Hyperthreading. Deathman, myself, evilsizer and many other old school contributing members on this forum have tested on/off HT for many games and apps. Most games will lose about 5% to 7% of FPS due to added latency from 4 extra threads not to mention as we have proven in this very thread, you will gain about 100 to 200 mhz on your overclock + lower temps by 10*C each core on average because of less strain on each core. Many newer games and apps are Multicore optimized for physical cores, but I would have to say there is less than a 5 percentile otpimized for multicore + Hyperthreading. So, to answer your question yes your amount of cores in windows task manager will go from 8 to 4, but those extra 4 cores/threads are for the most part adding latency and slowing down total processing time UNLESS the game/app is HT optimized. You will be very hard pressed to find any game currently on the market that is HT optimized.

Let me know if you have any other questions.

Dom

[saltytheseagull]

I just had the same problem with my X58 gigabyte / 6GB OCZ triple channel..

I'm experiencing a similar problem as well with X58UD5 and Mushkin Redlines. Everything in windows shows 6GB, but when I do each slot in CPU-Z only slots 1 and 3 show up. Also, in memtest only two of the sticks show their SPDs but it shows 6GB. Any thoughts?

[NGL_BrSH]

I'm having the same issues and i'm going out of my tree.

[NGL_BrSH]

i'm having the same problem.. returning it all to microcenter tomorrow and getting a new brand of ram and another UD5.

[MmmDieselFumes]

Alright, sorry for the long in-between-er again lol. I went shopping once I finally hit upon what I thought might be my issue with the ram. I noticed that one of my exhaust fans on the top of my case was actually barely touching the locks for the ram dimms, and I figured that maybe the vibrations of the fan were slowly making the dimms back out. So I took out that fan, ran it for a day or two, and so far, I haven't had any issues with the ram not showing up (keeping my fingers crossed).

So anyways, I went shopping today. I managed to score a brand new dell 22" full 1080p, 1920x1080, 5ms with HDMI monitor from best buy for $139.99 today. Probably one of the best scores I've every made at BB. They had it miss labeled, so I had the gf throw a god almighty horrendous screeching b**** fit, and they graciously honored the wrong card in front of the monitor.

I also happened to glance in their other aisle and I saw a single lonely Corsair H50 sitting there screaming my name with a sweet price tag of $59.99 (it's currently $79.99 @ newegg. Was going to order it today, but got a wild hair up my *** to go to BB), so I picked up that as well.

So I've spent the last few hours installing this new h50 and SQUEEZING (barely, had to use some zip ties in places) all 6 case fans into this TINY "mid tower" ThermalTake case lol. And it all looks killer with my wire routing "skills"/ocd. I'll have to take some pictures whenever I get the motivation to do so...

Anyhow, I turned back on HT, and put the juice back up to auto just to take a gander and see how well she preforms and cools now. Needless to say, with prime95 doing a 8-threaded torture blend test, I am pleasantly impressed with it's cooling over that horrible stock intel hs/f.

Picture:
http://i302.photobucket.com/albums/nn114/purecountryrebel/Capture-2.jpg

Note: Yeah I really dunno what to do with all this new found screen realestate! It's awesome!

And, I guess now's about time I start screwing with actually overclocking this rig eh?

[MmmDieselFumes]

Sweet baby badonk adonk! 4ghz with this h50! Awesome! Turned off HT again, turned vcore up to 1.3, 200x20 just like you told me Dom! And over 15mins on prime95, and it's still stable @~65ºc!

Yep, that made my day!

http://valid.canardpc.com/cache/banner/1136411.png

CPU-Z (http://valid.canardpc.com/show_oc.php?id=1136411)

http://i302.photobucket.com/albums/nn114/purecountryrebel/Capture-3.jpg

[MmmDieselFumes]

Now I've gotta figure out how to play with the ram speeds... But it's time for bed. If I don't drag my self away right now, I'm not gonna be going to work tomorrow... Lol

[MmmDieselFumes]

I couldn't help it, had to do one more thing...

Running 4.2ghz now (200x21, 1.3v), seems stable, temps oddly enough exactly where they where when running a multiplier of 20...

http://valid.canardpc.com/cache/banner/1136435.png (http://valid.canardpc.com/show_oc.php?id=1136435)

http://i302.photobucket.com/albums/nn114/purecountryrebel/Capture-4.jpg

[MmmDieselFumes]

And finally (yeah I know, I lied about going to bed... I will right after this lol)

4.2ghz @ 1.3v/vcore, 800mhz @ 1.5v/dram. Have no idea if that's a decent OC or not (failed to boot with mem at 1000mhz using that xmp deal), but I do remember not to move the dram and vcore more than 0.5v away from each other. But setting my ram from 600mhz to 800mhz and still being on the same 1.5v sounds OK to me (but then again, what do I know?)

Hoping someone will look at what I posted and give me a little guidance, whether good job, or constructive criticism. Gonna leave prime95 burning all night and check it in the morning.

[/rambling]

http://valid.canardpc.com/cache/banner/1136456.png (http://valid.canardpc.com/show_oc.php?id=1136456)

http://i302.photobucket.com/albums/nn114/purecountryrebel/Capture-5.jpg

[MmmDieselFumes]

And actually, I was able to bring that QPI back down to 1.30000 instead of that 1.37500 that's there in that pic, and it's still stable.

Done, for real now.

[baditude_df]

1.3v on the QPI/DRAM should be more than sufficient to run that bclock with 6GB of ram at 800MHz. You could probably drop it even more TBH, but that;s where I leave mine for 24/7 stability.
Really the only thing that you're pushing hard is your chip with that high multi, so other than making sure your Vcore is in line, the rest should be fine as is. vDIMM at 1.5v regardless of the manufacturer's specs is something that might need to be bumped up at some point for greater stability. Just keep an eye on it.
A 1.3v 4.2GHz with that processor is quite an achievement if you can keep it stable. Personally I'd just back it off to 4GHz and get the vcore lower, but that's because I run HT for Folding.

[mastrdrver]

Have you ever reseated your cpu? A lot of times the "4gb instead of 6gb" is because the cpu is not seated properly on the pins. VRef can also play a part in this too as some have come to find out.

Detecting less memory than what's installed (http://i4memory.com/f54/asus-x58-6-12gb-memory-overclocking-detecting-less-memory-than-installed-19200/)

[MmmDieselFumes]

It's still only showing me 4gb every now and again. And yeah, I've reset it a couple of times, always with the same issue. Thanks for the link though, that's some real good information.

I'm pretty sure my problem lies in the motherboard itself, probably some bad connections on the b1 module, since it's sporadic, and it will "fix" itself if you wiggle the stick of ram in the b1 slot. Since I've had this board for around 3 months now (I think...), I'm more than sure that I can no longer RMA it to newegg where I bought it, so I really don't know what I should do. If anyone has some advice on what I should do with this board (contact ASUS maybe? Tell them what?) I would be more than appreciative!

[Sherlock1234]

Well guys new to this forum on account of this exact issue. I have a p6tse I7 950, 2 60 GB OCZ SSDs, 500 GB SATA 5400 RPM, 12 GB ram. I purchased this from IBUYPOWER, this started happening initially. I started with 6 gb of ram. I would open the box and pull out the RAM and then put it back in. It would go from 4 gb to 6 gb when I did this. Now that I have doubled my RAM, it started at 12 GB ram and then now it says 8 GB ram. I fully believe that the problem with the computer is the motherboard. We all have different RAM but we all have the same motherboard. Personally I am fine with 8 GB of RAM but I am angry with the fact that my motherboard is not working properly.

[Neckername]

Well this is a very common problem, so before you go replacing dimms or buying a new cpu (if you think your memory controller is bad), do this:

1: go to start>run (search for "run" in vista/win7)> then type "msconfig" in run (without quotes)
2: go to the boot tab and click advanced options. Make sure the maximum memory checkbox is unchecked. If it is checked, Un-check the box and hit ok, then hit apply, and then hit ok and restart your system.
3: make sure you have the latest BIOS, if you do not, you can go to your motherboard manufacturers' website and find it in the support/drivers section.
If your BIOS is up to date skip to step 8.
4: Download the BIOS and burn it to a CD or Flash Drive.
5: restart the computer and insert the CD or flash drive you added the BIOS image on
6: set ALL BIOS SETTINGS to DEFAULT!
6: go into the boot settings right on the startup using the hotkey specified on startup, and change the first boot device to CD/DVD drive or USB (depending on what you used to store the BIOS image).
7: Flash the BIOS, set the settings back to normal and restart the computer. Usually the manufacturer of the motherboard will have their own BIOS flash procedures on the website and/or included in the BIOS image download.
8: Restart the computer and go into the BIOS settings.
9: Go into the memory options and go to TRL (this setting may be in the memory timings section of the options), and change the value to +2. This allows for the DIMMS to be initialized/detected before the handoff to the OS!
10: While in memory options set the CHSTRAP or MCHSTRAP to "DRAM RATIO"
11: Go into the voltage settings and set the RAM voltage to a supported value. Check on the box the modules came in or check with the manufacturer before you do this, as setting the voltage higher than the supported value could damage the DIMMS and/or void the warranty.
12: Save the settings (usually by hitting F10) and restart.


That should fix any memory initialization/detection issues. If this does not fix your issue. Run memtest86 to see if one of the DIMMs/slots is bad. If all comes back ok and you still have problems, there could be a slight chance, that for some reason, your memory controller inside the CPU is on the fritz, meaning you may have pushed memory and/or CPU clocks and/or voltages too far, and you may need a new CPU. However, that is rarely the case. Hope this helps!

[Neckername]

also check to see if memory remapping is an option in the BIOS memory settings. If so, enable it and that should help as well.