CMX4GX3M2A1600C9 BSODs

Hi all,

 

I built a new desktop this summer and have had frequent BSOD problems ever since.

About 95% of the blue screens have the same STOP code and parameters:

 

MEMORY MANAGEMENT

0x0000001a (0x0000000000041790, 0xfffffa8001556a60, 0x000000000000ffff, 0x0000000000000000)

 

but very infrequently I will receive others, e.g.:

 

0x0000001a (0x0000000000041284, 0x0000000005473001, 0x00000000000020c2, 0xfffff70001080000)

0x0000001a (0x0000000000005002, 0xfffff70001080000, 0x00000000000011e2, 0x020011e3fffffffe)

0x0000001a (0x0000000000000031, 0xfffffa80058498a0, 0xfffff880038bd000, 0xfffff8a003c2eaab)

 

and just today I had the excitement of my very first non-MEMORY MANAGEMENT STOP:

 

NTFS FILE SYSTEM

0x00000024 (0x00000000000c08a2, 0x0000000000000000, 0x0000000000000000, 0x0000000000000000)

 

But again, almost every single BSOD (and I've had dozens of them) has been MEMORY MANAGEMENT, and specifically, has had a parameter1 value of 0x41790. Hence I think it is fair to say that my problems are, in some way, memory-related.

 

The blue screens have only popped up when I am running with both sticks installed at once. I can't prove that they don't happen with only one stick in at a time, but I can definitely say they must be much, much less frequent, because I have not experienced a single one yet. By contrast, with both sticks in, I can have blue screens every five minutes on a bad day.

 

I should also mention that this is actually my replacement set of memory - I RMA'd my first pair (same model) with NewEgg because it was having the same problems. I have tested each stick in each slot with memtest, with conflicting results. Generally they run fine, even for 10+ passes overnight, but occasionally I will find errors.

 

I have tried adjusting the timing and voltage settings per CPU-Z, but this has not reduced the frequency of the blue screens at all. I have tried the mobo's default settings, I have tried 1333 MHz 9-9-9-24-34-1T at 1.50V, I have tried 1600 MHz 9-9-9-24-41-2T at 1.64V (my BIOS does not give 1.65V as an option, only 1.64 and 1.66V), and the problems persisted.

 

I have downloaded the latest drivers for my graphics card, the latest firmware for my SSD, the latest everything from Gigabyte, and nothing seems to fix the problem except running with only one stick in at a time. As I paid for 4GB dual-channel RAM, this is not a viable solution.

 

The NTFS FILE SYSTEM error today was rather odd, and makes me wonder if it is not SSD-related. I ran chkdsk /r and quite a few errors were found and fixed. This did not, of course, fix the blue screen problem. I tend to doubt if the SSD really is the culprit, because I have read too many other threads of people having this issue and not all of them use an SSD.

 

I feel like the most likely culprit at this point is probably bad DIMM slots. I am very, very unhappy about the prospect of returning my motherboard and would like to reserve this as a last resort. Can anyone think of any other possibilities?

Bad DIMM / reseat the CPU (possible bent pin).

Wired,

 

What makes you suspect the CPU?

Memory controller's in the CPU, and one bent pin (if it's one that the memory controller uses) can cause issues that seem memory related.

About 95% of the blue screens have the same STOP code and parameters:

 

MEMORY MANAGEMENT

0x0000001a (0x0000000000041790, 0xfffffa8001556a60, 0x000000000000ffff, 0x0000000000000000)

 

Sorry to bump this but I had the same problem and it was due that the RAMs are not installed in their Dual Channel configuration. When corrected, the issue immediately disappears.

 

For example, on the ASUS Crosshair IV Formula, RAMs installed in Dual Channel configuration are placed in the RED or BLACK slots respectively. Not one in red and other in black.

Sorry to bump this but I had the same problem and it was due that the RAMs are not installed in their Dual Channel configuration. When corrected, the issue immediately disappears.

 

For example, on the ASUS Crosshair IV Formula, RAMs installed in Dual Channel configuration are placed in the RED or BLACK slots respectively. Not one in red and other in black.

 

Thanks for the tip, fireblaster, and I appreciate the bump. Unfortunately for me, I have checked, re-checked, and re-re-checked the manual for the GA-P55A-UD3P's DIMM slots and the grey slots are the two that must be populated first; the blue slots are to be populated second. I am positive that I have the RAM in the correct slots.

 

I should also add an update to the thread as long as I am here. Still with only the one stick of RAM installed, I switched my computer off for the weekend as I was going out of town. When I got back in last night, the computer failed to load Windows, repeatedly. Win7 attempted "startup repair," something that must be new to the version, but this also resulted in a blue screen after about 30 seconds. I tried swapping the one stick for another, but this did not fix the problem. I left "startup repair" going overnight, and still going while I was at work, but it wasn't getting anywhere, so I moved the stick to the other DIMM slot and--for now--the computer is working again.

 

I left my tubes of thermal compound at my parents' house and have not had a chance to head out there recently to try to reseat the CPU and check for bent pins or other problems. I will post an update when I do so.

 

By the way, the errors I got last night were rather exciting, since only some of them were MEMORY MANAGEMENT. I also got:

 

0xc000021A {Fatal System Error}

0xc000001D (0xFD241607, 0x0070F180)

0x0000007E {0Xffffffffc0000006, ......}

 

Aren't $1500 brand-new computers fun?

...recently to try to reseat the CPU and check for bent pins or other problems. I will post an update when I do so.

 

Aren't $1500 brand-new computers fun?

 

Please do post an update. (:

The CPU could be your ticket to getting the $1.5k computer working. With RAMs swapped, the only solution left would be the CPU itself.

 

All the best!

Hi Spum,

 

I have the same RAM as you (but GA-P55-UD4 and i5-750). I recall having some initial BSODs like yours, but these settings cured them all (in case you haven't already tried):

 

1. In the M.I.T. menu, submenu Adv. mem.settings: Enable XMP to Profile1

leaving System Memory Multiplier and DRAM Timing Selectable on Auto.

2. In submenu Adv.volt.settings: DRAM voltage from 1.5V to 1.64V and QPI/Vtt to 1.15V.

 

I hope that the BIOS menu selections correspond to your m/b version and that the voltage settings are relevent for your CPU.

UPDATE: Shortly after my last post, problems got a lot worse. The non-MEMORY MANAGEMENT blue screens started happening like crazy, and I decided to try the "Startup Recovery" feature which must be new to either 7 or Vista. It found no fixable problems, yet it somehow managed to get itself into an infinite loop. From what little I could find online on my phone about this problem, there is no known way to get out of the loop.

 

So, I reformatted and reinstalled Windows. I also stopped by my parents' house to pick up my thermal compound, and totally reassembled the computer, including re-seating the CPU. I did not see any bent or broken pins.

 

At this point, I am leaning towards a faulty motherboard, probably the DIMM sockets. It could just be a bad CPU, as well.

 

Hi Spum,

 

I have the same RAM as you (but GA-P55-UD4 and i5-750). I recall having some initial BSODs like yours, but these settings cured them all (in case you haven't already tried):

 

1. In the M.I.T. menu, submenu Adv. mem.settings: Enable XMP to Profile1

leaving System Memory Multiplier and DRAM Timing Selectable on Auto.

2. In submenu Adv.volt.settings: DRAM voltage from 1.5V to 1.64V and QPI/Vtt to 1.15V.

 

I hope that the BIOS menu selections correspond to your m/b version and that the voltage settings are relevent for your CPU.

 

Cooldrive, thank you very much for the suggestion. Yes, our motherboards share the same BIOS, and yes, the i7-860 is also a 1.65V processor. I have tried Vcore = 1.64 for some time, but I have never changed the QPI. I'll let you know what happens. My "stress test" is to put both sticks of RAM in and boot to Windows!

UPDATE:

 

As I mentioned last time, I totally disassembled the computer, reformatted and reinstalled Windows, and adjusted the memory controller voltage per Cooldrive's recommendation. All seemed well for a day or two, as I was able to boot with both sticks of RAM in, but now I am back to having regular blue screens if I try to use both. It's working okay now with one stick, but if the past is anything to go by, it will only get worse over time. I have tried other QPI voltages, but they do not seem to resolve the problem. I have tried downclocking the RAM all the way to 800 MHz, I've tried adjusting the DRAM voltage...nothing seems to work.

 

Since it would be far easier for me to adjust BIOS settings than replace my motherboard (without even knowing for sure that that is the problem), does anyone know of a complete set of working BIOS settings for my combination of RAM and processor?

In case it helps anyone, here is a dumpchk.exe analysis of the most recent crash:

 

*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

Use !analyze -v to get detailed debugging information.

BugCheck 1A, {41790, fffffa8000f9da10, ffff, 0}

Probably caused by : ntkrnlmp.exe ( nt! ?? ::FNODOBFM::`string'+33906 )

Followup: MachineOwner
---------

2: kd> !analyze -v
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

MEMORY_MANAGEMENT (1a)
   # Any other values for parameter 1 must be individually examined.
Arguments:
Arg1: 0000000000041790, The subtype of the bugcheck.
Arg2: fffffa8000f9da10
Arg3: 000000000000ffff
Arg4: 0000000000000000

Debugging Details:
------------------


BUGCHECK_STR:  0x1a_41790

CUSTOMER_CRASH_COUNT:  1

DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT

PROCESS_NAME:  WerFault.exe

CURRENT_IRQL:  0

LAST_CONTROL_TRANSFER:  from fffff80002f2df9e to fffff80002eba740

STACK_TEXT:  
fffff880`03c88828 fffff800`02f2df9e : 00000000`0000001a 00000000`00041790 fffffa80`00f9da10 00000000`0000ffff : nt!KeBugCheckEx
fffff880`03c88830 fffff800`02eeddf9 : 00000000`00000000 000007fe`f8d26fff fffffa80`00000000 fffffa80`00000000 : nt! ?? ::FNODOBFM::`string'+0x33906
fffff880`03c889f0 fffff800`031d31d0 : fffffa80`0622bb90 0007ffff`00000000 00000000`00000000 00000000`00000000 : nt!MiRemoveMappedView+0xd9
fffff880`03c88b10 fffff800`031d35db : 00000000`00000000 000007fe`f8b70000 fffffa80`00000001 fffffa80`064bf010 : nt!MiUnmapViewOfSection+0x1b0
fffff880`03c88bd0 fffff800`02eb9993 : fffffa80`05c121e0 fffff880`03c88ca0 fffffa80`068b9b30 00000000`00436fbc : nt!NtUnmapViewOfSection+0x5f
fffff880`03c88c20 00000000`7740fffa : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiSystemServiceCopyEnd+0x13
00000000`001ebde8 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : 0x7740fffa


STACK_COMMAND:  kb

FOLLOWUP_IP: 
nt! ?? ::FNODOBFM::`string'+33906
fffff800`02f2df9e cc              int     3

SYMBOL_STACK_INDEX:  1

SYMBOL_NAME:  nt! ?? ::FNODOBFM::`string'+33906

FOLLOWUP_NAME:  MachineOwner

MODULE_NAME: nt

IMAGE_NAME:  ntkrnlmp.exe

DEBUG_FLR_IMAGE_TIMESTAMP:  4c1c44a9

FAILURE_BUCKET_ID:  X64_0x1a_41790_nt!_??_::FNODOBFM::_string_+33906

BUCKET_ID:  X64_0x1a_41790_nt!_??_::FNODOBFM::_string_+33906

Followup: MachineOwner
---------

Hi again Spum,

 

I regret your seemingly endless troubles with your PC. It's an obvious suggestion to ask you to hit your m/b with an axe to end your suffering permanently.

Seriously, my impression is that your are facing a Bios that is f***ed up in some way. Since you have had your PC up and running, for a few days anyway, with OK behaviour, it doesn't smell faulty h/w. BSODs, frequent restarts, etc, could be a sign of Bios autoconfiguring itself after each restart, creating correct values that are contradictory to other values (previous set manually by you),

You have tried every setting possible in the Bios, but nevertheless I suggest that you you start all over from scratch in a super structural way. If you have done this a thousand times before or if you consider my text below to be too elementary, just forget it.

Start by clearing CMOS and load optimized defaults. Change only what's necessary to continue to work. Then enter the settings for DRAM timings/latency (if necessary), voltage, XMP profile and Qpi/Vtt voltage. Also, temporarily disable Intel Turbo Boost, all settings concerning CPU thermal/stepping settings and disable USB Storage function and USB Legacy function. After each restart, make a note of what you changed and the corresponding result.

If you still experience the same BSODs as before, my only remaining suggestion would be to buy a pair of cheapest possible compatible 1 GB RAM modules and test them instead of your existing DRAM. If I'm allowed to suggest it, maybe it would be a good idea to try another brand (for pure research purposes only, of course). Since you''re running Win64-bit , you could eventually use the additional RAM if you get your PC to work. If that also turns out to be a dead end, I see no other way than to replase m/b/CPU or consider my initial suggestion.

I wish you the best of luck in the BSOD-fog.

Hi again Spum,

 

I regret your seemingly endless troubles with your PC. It's an obvious suggestion to ask you to hit your m/b with an axe to end your suffering permanently.

Seriously, my impression is that your are facing a Bios that is f***ed up in some way. Since you have had your PC up and running, for a few days anyway, with OK behaviour, it doesn't smell faulty h/w. BSODs, frequent restarts, etc, could be a sign of Bios autoconfiguring itself after each restart, creating correct values that are contradictory to other values (previous set manually by you),

You have tried every setting possible in the Bios, but nevertheless I suggest that you you start all over from scratch in a super structural way. If you have done this a thousand times before or if you consider my text below to be too elementary, just forget it.

Start by clearing CMOS and load optimized defaults. Change only what's necessary to continue to work. Then enter the settings for DRAM timings/latency (if necessary), voltage, XMP profile and Qpi/Vtt voltage. Also, temporarily disable Intel Turbo Boost, all settings concerning CPU thermal/stepping settings and disable USB Storage function and USB Legacy function. After each restart, make a note of what you changed and the corresponding result.

If you still experience the same BSODs as before, my only remaining suggestion would be to buy a pair of cheapest possible compatible 1 GB RAM modules and test them instead of your existing DRAM. If I'm allowed to suggest it, maybe it would be a good idea to try another brand (for pure research purposes only, of course). Since you''re running Win64-bit , you could eventually use the additional RAM if you get your PC to work. If that also turns out to be a dead end, I see no other way than to replase m/b/CPU or consider my initial suggestion.

I wish you the best of luck in the BSOD-fog.

 

As an update, I have been booting from a Linux CD for a couple of days now with no blue screens (or their Linux equivalents). I had hoped that this test would identify whether the problem was due to a faulty motherboard or some conflict due to Windows. However, now I am not so sure that booting from a CD is equivalent to booting to the HD. Although the prospect doesn't thrill me, I should probably install XP or Linux to the SSD and try that for a few days. Speaking of which, I ran a Seagate hard drive diagnostic (******** does not offer one), and my SSD passed. Hence, although I don't know if I am on firm ground here, I am leaning towards there being a BIOS/Windows problem. In case I didn't yet mention it, the problem cannot be due to a third-party driver, since Windows crashes in safe mode (even without networking).

 

Cooldrive, I agree that I have probably tried all I can without trashing the motherboard or memory, but your idea is worth a shot. I'll see what happens. I appreciate your continuing advice.

I did exactly the same thing you did but with an Asus motherboard. I installed the OS, Win 7 64 bit and it ran for a day or so. It then kept getting a multitude of memory manager errors, NTFS errors, windows explorer errors, etc. I have low level formatted the HDD several times and changed the keyboard and mouse combo, the vid card and ram to this Asus recommended model. I am also at my wits end. I did notice however that the bios was reporting the ram as 1333 MHZ instead of its 1600 speed. Asus has told me to get the proper ram settings to manually set the voltage and timing. They say that will solve the ram problems. I am currently waiting for Corsair technical support to provide me with the proper settings. Then the ball is back in Asus's court. This is my first i series processor and I cannot definitely say that the problem is not solely with the ram since the memory management modules are now in the processor. I am hoping that improper ram settings are creating memory addressing issues with false results due to the improper settings. All I can do is wait for tech support. I look forward to your logs.

UPDATE: As expected, nothing I change in the BIOS seems to have any effect on the issues I have been experiencing.

 

I did, however, realize something that is rather peculiar: once my computer is up and running trouble-free, I can run it all day, for days on end, and even restart repeatedly, without problems. It is only when I shut off the computer for the night and do a cold boot in the morning that I have these problems. Googling "cold boot bsod" (no quotes) or even "cold boot 41790" with my specific error code, I see there are many, many people having this issue with Windows 7. I find it strange that everything works--memtest, prime95 under full load, music, hd video, reboots--when the computer is hot, but the system comes to a screeching halt when I try to boot cold. Does this suggest aspect of the problem a new line of attack?

I did exactly the same thing you did but with an Asus motherboard. I installed the OS, Win 7 64 bit and it ran for a day or so. It then kept getting a multitude of memory manager errors, NTFS errors, windows explorer errors, etc. I have low level formatted the HDD several times and changed the keyboard and mouse combo, the vid card and ram to this Asus recommended model. I am also at my wits end. I did notice however that the bios was reporting the ram as 1333 MHZ instead of its 1600 speed. Asus has told me to get the proper ram settings to manually set the voltage and timing. They say that will solve the ram problems. I am currently waiting for Corsair technical support to provide me with the proper settings. Then the ball is back in Asus's court. This is my first i series processor and I cannot definitely say that the problem is not solely with the ram since the memory management modules are now in the processor. I am hoping that improper ram settings are creating memory addressing issues with false results due to the improper settings. All I can do is wait for tech support. I look forward to your logs.

 

Spark, the proper voltage settings and timings are:

 

V_DRAM = 1.65 V (use 1.64 if not available)

Timings = 9-9-9-24-41-2T

Frequency = 12x (i.e. 1600 MHz)

 

I believe that Intel does not officially support any memory frequency above 1333 MHz, and hence even 1600 MHz-labeled memory is considered to be "overclocking" the CPU's memory controller. Since 1600 is only negligibly faster than 1333, you might choose instead to use:

 

V_DRAM = 1.50

Timings = 9-9-9-24-34-1T

Frequency = 10x (i.e. 1333 MHz)

Hi again, Spum

 

I understand that your PC is still haunted by BSODs. How about some alchemy (in case you haven't tested it yet. Or maybe the settings below are already made automatically for you by your CPU/Bios):

Anyway, if you're still running your CPU with BCLK 133 and a bus/core ratio of 21, then take a chance and change BCLK to 166 with a ratio of 17. I borrowed this "advice" from an OC Forum and evidently it rescued someone there from going mad.

Hi again, Spum

 

I understand that your PC is still haunted by BSODs. How about some alchemy (in case you haven't tested it yet. Or maybe the settings below are already made automatically for you by your CPU/Bios):

Anyway, if you're still running your CPU with BCLK 133 and a bus/core ratio of 21, then take a chance and change BCLK to 166 with a ratio of 17. I borrowed this "advice" from an OC Forum and evidently it rescued someone there from going mad.

 

No luck. This severely overheated my CPU (>100F) and caused it to shut down and fail to boot until I removed the motherboard's battery. Of course, I'm using Intel's miserable stock cooler, so this result may not be typical, but that's still a huge temperature rise! For that reason, I would not recommend these settings to anyone else who might be having these problems and reading this thread.

 

I tried all base clocks between 133 and 166 which are integer divisors of ~2800, but with no luck booting with any. Only 133 x 21 seems to work.

Spum, don't be too quick to conclude that this is a Win7 or Intel problem. I've had plenty of similar problems with an ASUS mobo, AMD Phenom II, win XP64, and theTW3X4G1333C9a kit, including the greater instability with both sticks compared to just one and the cold vs. warm boot issue. My system is now starting reliably and running stably with a CMV4GX3M2A1333C9A kit, with default BIOS settings. Based on my very limited experience (n=1) and no engineering know-how at all, the post suggesting you try other claimed compatible memory is a good one-it could be that finding a mobo-DDR3 match is much more chancy than the guarantees of compatibility would suggest. Bon chance!

Spum, don't be too quick to conclude that this is a Win7 or Intel problem. I've had plenty of similar problems with an ASUS mobo, AMD Phenom II, win XP64, and theTW3X4G1333C9a kit, including the greater instability with both sticks compared to just one and the cold vs. warm boot issue. My system is now starting reliably and running stably with a CMV4GX3M2A1333C9A kit, with default BIOS settings. Based on my very limited experience (n=1) and no engineering know-how at all, the post suggesting you try other claimed compatible memory is a good one-it could be that finding a mobo-DDR3 match is much more chancy than the guarantees of compatibility would suggest. Bon chance!

 

Well, we have a winner. I traded RAM with someone else (a pair of Corsair CM3X2G1333C9D 2GB sticks), and presto! No problems for either one of us. I might add that the heat sinks on the new sticks look pretty slick.

 

Since I now know that it is a compatibility issue, I Googled my RAM and mobo combination, and, surprisingly, found several people running it just fine. Not sure why I had problems with not just one pair, but two, of this memory. For anyone reading this thread down the line, I still can't recommend the combination of Corsair's CMX4GX3M2A1600C9 memory with Gigabyte's GA-P55A-UD3P motherboard. There's nothing wrong with either component, but together, Windows 7 (the 64-bit edition, at least) was not stable for me at any voltage or frequency settings. If you're getting MEMORY MANAGEMENT blue screens with error code 0x41790, this is your problem.

 

Of course, you never know when a BSOD will strike next. I'll be back if the problems return. Many thanks to everyone who helped me pin down the cause!