Is my mobo screwed up?

[Slysoft]

So I've had some issues ever since I upgraded my computer 7 months ago. At first I was getting random BSOD which after running memtest86 I found a bad stick of ram and pulled it, and my PC began running fine. A few months later though, my PC started hard freezing randomly (everything locks up, USB devices lose power, static sound coming from speakers). This would happen randomly whether I was doing anything or not and while infrequent was still a problem. Recently, my PC has begun simply turning off randomly and not posting on reboot, with a post code that isn't in the manual and the customer service people don't know what it is.

Ultimately, I managed to get the system to post by removing all of my ram except for one stick. I'm not sure if the problem is fixed but it hasn't happened yet so I'll have to keep monitoring, but one thing to note is that I've tried 2 different sticks in the 2nd ram slot and the computer will not post until I remove it. I haven't tried the ram in the 3 or 4 slot yet but I have a feeling that both sticks can't be bad, so I'm thinking its a fried ram slot. I hope not because that will put me out of commission for a week or two.

[halfelite]

Did you check the ram is compatible with the board? It might need more or less power then default settings is giving.

[krumm]

Is it Intel or AMD.  If AMD it could be the cpu, due to the memory controller being on it.  I had a bad video card cause similar issues as well.  from what you have said I'm guessing it's Intel and the memory controller on the mobo is bad.  See if the computer posts with the ram only in the 1st ram channel, and then only in the 2ed ram channel.  look in the manual if you don't know which slots that is.  Some mobos require ram in certain slots( the manual should tell you if that's the case), but if not then you can maybe find out if it is the memory controller is only working on one channel.  It may be pointless to even go through with diagnosing and just replace the mobo out right.

I would not wait to long tho.

I had a CPU at work that had a bad ram controller and did not work with ram in the first channel, so I just used the 2ed channel until it died completely like 2 weeks later, so time may not really on your side.

This is all based on me assuming it's the ram controller causing the issues(if it fits the issue it usually is).


Incompatible RAM should have always had the same symptoms and not change.

[nstgc]

Modern Intel CPUs also have an integrated memory controller.

Slysoft, could you provide some more information such as your motherboard, RAM, CPU, and  PSU?

Also, it could really be anything. For instance I found that my HDDs were incapatible with the controller. My Samsung F3 (Not Seagate Samsung) didn't play nice with my ICH10R (which is integrated onto my Intel CPU). Computer wouldn't POST and crashed at random. Replaced them with Western Digital RE4s and haven't have any problems. I thought it was my RAM, but it turned out to be the HDDs.

It still is mostly likely either your RAM or PSU, but it could be anything.

[Slysoft]

Yeah, I haven't tried using this one stick in all slots yet, but that was the next thing I was going to do. It is an intel CPU.

Mobo
Ram
PSU
CPU

[nstgc]

Chances are your RAM is compatible with that mother board. There isn't a compatibility list on the EVGA product page, but I'd bet money on it. The PSU should be good, and you're CPU does have an integrated memory controller.

I would first check with EVGA's tech support to see if there is a compatibility issue, at the same time I'd discuss it with Corsair. They may ask that you RMA your RAM or PSU.

[rostheferret]

...my PC started hard freezing randomly (everything locks up, USB devices lose power, static sound coming from speakers). This would happen randomly whether I was doing anything or not and while infrequent was still a problem.

Ultimately, I managed to get the system to post by removing all of my ram except for one stick. I'm not sure if the problem is fixed but it hasn't happened yet so I'll have to keep monitoring, but one thing to note is that I've tried 2 different sticks in the 2nd ram slot and the computer will not post until I remove it. I haven't tried the ram in the 3 or 4 slot yet but I have a feeling that both sticks can't be bad, so I'm thinking its a fried ram slot. I hope not because that will put me out of commission for a week or two.

Firstly, most RAM sometimes messes up if you put it in the wrong slot. Sometimes it's colour coded, but usually the 1st/3rd are linked, as are the 2nd/4th, and it'll look to the 1st slot initially. Since you have 6 slots, I wonder if it's not the 1st/3rd/5th that are linked, just try them out. I'm sure someone can elaborate better as to why all this is. Another thing to check is that the BIOS is clocking them at the right speed. The RAM's rated at 1600mhz, make sure it's actually running that from the BIOS.

As far as the freezing is concerned, I had the same problem when I first built my PC - I even sent it to a PC repair shop who couldn't figure it out - and eventually I stumbled into what was happening. A lot of mobo's these days have 'green' functions which limits power to devices that aren't actively in use, and some devices - in my case the graphics card - don't exactly take kindly to having it's power yanked off. A BIOS update fixed the problem so it'd be worth trying that.

Sadly, shit like this really COULD be anything. It definitely sounds like a hardware issue though, best suggestion is to unplug anything non-essential; graphics card, sound card, non-OS HDD (even the OS HDD if you can RAID/install an OS to another disk), then plug it all back in one at a time until you know exactly what's causing the issue.

[halfelite]

Chances are your RAM is compatible with that mother board. There isn't a compatibility list on the EVGA product page, but I'd bet money on it. The PSU should be good, and you're CPU does have an integrated memory controller.

I would first check with EVGA's tech support to see if there is a compatibility issue, at the same time I'd discuss it with Corsair. They may ask that you RMA your RAM or PSU.

EVGA hides there stuff here is there list for anyone else

http://www.evga.com/support/motherboard/

and his ram is not listed

But i would also try on the evga forums they are usually good at helping people out and a couple evga employees also read posts there

[Freedom Kira]

Firstly, most RAM sometimes messes up if you put it in the wrong slot. Sometimes it's colour coded, but usually the 1st/3rd are linked, as are the 2nd/4th, and it'll look to the 1st slot initially. Since you have 6 slots, I wonder if it's not the 1st/3rd/5th that are linked, just try them out. I'm sure someone can elaborate better as to why all this is. Another thing to check is that the BIOS is clocking them at the right speed. The RAM's rated at 1600mhz, make sure it's actually running that from the BIOS.

It depends. From a past thread on here, it seems that AMD boards have each channel grouped together (first two are one channel, last two are second channel), while Intel's are striped the way you said (alternating channels).

Anyway, the symptoms OP are describing are reminiscent of blown and/or leaking capacitor problems I've had with older boards. Might this be the case, perchance? It's a relatively new board so I really doubt this could be the issue, but you never know.

[limefc]

That board has all solid capacitors. There is no possibility of "leaking capacitor" and in the end, leaking capacitor is one of the possible causes of the real problem anyway, which is known as a fucked VRM. Which is a possibility even today - A ruined capacitor alone on a motherboard will not necessarily cause a computer to malfunction. I had an old computer with no intact capacitors left on the motherboard and it was stable as a rock.

At first I was getting random BSOD which after running memtest86 I found a bad stick of ram and pulled it, and my PC began running fine.

Assuming this is the truth, the typical possibilities are somewhat limited. Either the motherboard is a dud and needs to be replaced or the power supply is royally fucked and delivering out of spec power, in which case the motherboard is probably already damaged anyway thus still in need of replacement.

The way I see it, there is no good news to be found here. I mean if your shit is freezing and it's not because of a botched overclock, then you're not looking at a good situation no matter how you slice it.

[nstgc]

Ultimately, all these comments boil down to "contact tech support". Check with EVGA to see if the RAMs absence from their approved list is an issue (some times its just not listed). Check with Corsair to see if they are aware of any known issues. If both give you the green light RMA the board. If not, get some new RAM that is on the list. If that still doens't work, I'm guess its the PSU, in which case you may want to try to get an RMA for that.

[limefc]

RAM compatibility lists are checklists of tested parts. RAM on the list basically means that this RAM was tested with the board, JEDEC and/or XMP profile was properly loaded and the RAM was autoconfigured.

Basically it means fuckin' nothing. Would be useful if it was comprehensible and only contained RAM that didn't work but alas nobody has the time to test all the different RAM in the world.

[nstgc]

I am surprised that the Vengeance wasn't on their. I would think that EVGA would test all "gamer" RAM, which the Vengeance falls into. That does suggest that it didn't make it, but it also could be that they simply didn't test it.

[limefc]

The "Vengeance" RAM is XMS3 with obnoxious heatsinks.

[nstgc]

The "Vengeance" RAM is slower XMS3 with obnoxious heatsinks.

fixed.

I have their XMS3 RAM, its tested at 2000MHz and I can verify that it works at speeds of at least 2166 (or whatever it is) without loosening the timing.

[limefc]

Performance wise, the difference between CLx at 2166 and CLx at 2000 is very little. I'd be surprised if RAM tested at 2000 didn't do 2166 without touching the timings. Not as much if it was Corsair XMS3 I guess, but still. It's like a 0.7ns difference. For comparisons a lot of cheapo RAM will still boot with an overclock of 1.25ns. Some better examples will actually be stable.

My Corsair XMS3 trash just barely booted with a 1.25ns overclock. However, my current Kingston cheap-ass RAM I managed to get to boot with a 2.25ns overclock and with some serious volts I was getting close to stability. 400+mhz overclock without touching timings or voltage? More likely than you think.

I'm digressing. Anyway, I'm pretty sure it comes from the exact same bin as XMS3 so it should be compatible with boards that are tested with XMS3 as working.

[nstgc]

The difference would likely be in the QC. The vengeance probably just didn't pass as many tests. Why they put those God awful heat spreaders on them, though, is a mystery.

As for the speed, I was just pointing out a difference. Its been shown that your computers performance with 1333 RAM is not much worse than 1600 and the difference between 1600 and 1833 is miniscule at best. I keep mine at 1833.

[limefc]

They are both tested the same way actually. Corsair takes the cream of the crop for their Dominator line of memory, anything that doesn't pass quality requirements for that will be sold as Vengeance, XMS3 and Corsair Value depending on heatsink (and bin).
Don't think they sell XMS3 DHX anymore, but that used to come from same bin as well and that's what I had.

1.8V XMS3 DHX. Sick. Voltage didn't do shit to them though, seriously. I could zap em with 2.1V for nothing or undervolt to 1.5V and still no effect.
Also memory frequency is deceptive.

1333MHz memory can be FASTER than 1600MHz memory. As in benchmark better.