Is my RAM dying or simple "wear"?

[Derek12]

Some time ago, I could OC my RAM to DDR2-800 and do some benchmarking with MAXXMEM and was fine, but after the benchmark I would clock to stock (DDR2-667) again, so I never used the RAM OCed except to benchmark as commented. The computer worked perfectly without hiccups when the RAM was OCed.


Now I wanted to bench it and clocked it to DDR2-800 again and after starting Windows, lots of error windows popped out, programs crashing, assertion faults(?) drivers (sound) not loading, DLL library errors, etc. Prime95 failed instantly, and got a single BSOD (CACHE_MANAGER) when rebooting.

Reverted to stock and all is fine again. But I underclocked to DDR2-400 in hope to "slow-down" the dying process if it's the case.

My question is, is my RAM dying or it's simply a bit "worn"? Could I use the RAM at stock as normally or should I prepare my wallet to buy another RAM?

Here is a SS of CPUZ, alongside Prime95 failing.




Many thanks

EDIT: I forgotten to mention, after restarting the computer, the POST halted and said entered to safe mode because a error or something but happened once.

[cadaveca]

Could be ram, could be memory controller, could be you updated BIOS since then?

[brandonwh64]

Whats your current ram voltage, and NB voltage?>

[Derek12]

Memory controller? that would be the CPU right? I hope not!!

Yes I updated BIOS some time ago but that was couple of years before even I OCed RAM to 800 first time

Originally Posted by brandonwh64

Whats your current ram voltage, and NB voltage?>

My RAM is currently at 1.872 and the Northbridge nope, because there isn't sensors nor a option to change it


I saw an option to upper RAM voltage, should I try it and see or could do some damage?

Thanks

[brandonwh64]

So the board does not have any voltages adjustments?

[Derek12]

Originally Posted by brandonwh64

So the board does not have any voltages adjustments?

According to BIOS only for RAM and CPU, I will try to up +0.1 to RAM and then:

If it still fails, could be the CPU or mobo?
If it works, could be the RAM is degraded?

Many thanks.

[cadaveca]

Could be the ram, for sure. Yes, I would up ram voltage and see if that fixes the problem..if it does, for sure it's the ram degrading.

[HalfAHertz]

You know it could be something as simple as a bad electrical connection(take the sticks out, get a pencil eraser and rub the connectors, then blow into the ram slots) or some dust bunnies that decided to cuddle up b/w the ram sticks...

[Derek12]

Originally Posted by cadaveca

Could be the ram, for sure. Yes, I would up ram voltage and see if that fixes the problem..if it does, for sure it's the ram degrading.

OK, I clocked it to DDR2-800 and upped voltage to 1.9 (sensor reading was 1.985) and all was fine, none errors, and Prime95 was fine aswell. Then we can conclude it's the RAM.


Is there an explanation why now I need to up the voltage to reach DDR2-800? Before it wasn't needed. So:

• My RAM is dying
  
• It worn out.

So I can use it normally or could I expect a RAM death eventually at medium term?

Many thanks


EDIT:

Originally Posted by HalfAHertz

You know it could be something as simple as a bad electrical connection(take the sticks out, get a pencil eraser and rub the connectors, then blow into the ram slots) or some dust bunnies that decided to cuddle up b/w the ram sticks...

I didn't thought that!. I will do it and I will update! also I will try to change slot

[cadaveca]

The ram may not be dying or degrading. Ram can be "burned-in" in such a way that it is not too happy with other timings and speed. You'll find many ram makers laud their "Burn-in technology", so it would be perfectly normal for something liek this to happen.


Bck in DD1 and DDr2 days, that's part of what we did searching for maximum benchmark scores. I had a rig set-up that was dedicated to running ram at the brink of stability, as it was found that doing so, over time, could make ram that was previously unstable...stable.

Originally Posted by Neuromancer

It could also be PSU or motherboard related if they are as old as the system itself, Capacitors might be wearing out and you need a higher voltage to provide same power to the sticks.

Definitely.

[Neuromancer]

It could also be PSU or motherboard related if they are as old as the system itself, Capacitors might be wearing out and you need a higher voltage to provide same power to the sticks.

IF you can borrow another set of DDR from someone to test and see if it needs a voltage increase to run at stock that would be the giveaway.

[Derek12]

Originally Posted by cadaveca

The ram may not be dying or degrading. Ram can be "burned-in" in such a way that it is not too happy with other timings and speed. You'll find many ram makers laud their "Burn-in technology", so it would be perfectly normal for something liek this to happen.


Bck in DD1 and DDr2 days, that's part of what we did searching for maximum benchmark scores. I had a rig set-up that was dedicated to running ram at the brink of stability, as it was found that doing so, over time, could make ram that was previously unstable...stable.

I didn't knew that thanks! Then does RAM can change the stability at any time if it's running out of spec without meaning that it's already damaged or degrading, This would make me happy

Originally Posted by Neuromancer

It could also be PSU or motherboard related if they are as old as the system itself, Capacitors might be wearing out and you need a higher voltage to provide same power to the sticks.

IF you can borrow another set of DDR from someone to test and see if it needs a voltage increase to run at stock that would be the giveaway.

Good point! I forget I had previous PSU incidents that killed various video cards and now I am using a new generic one until the other is repaired), could it affected the RAM? if so, then could it work as now or mean that it will pass away or degrade more and more, until it eventually dies? I measured the PSU voltages and are fine as for now!

Many thanks



I tried removing, cleaning the contacts, and changing the slot but no avail, if I don't up the voltage it's totally unstable (now Windows doesn't boot without BSODing) so I left it at stock

[Derek12]

Now crashes even at stock clock DDR2-667. Didn't BSODed yet, but some programs like Catalyst, Opera, Chrome and Winamp are crashing with the usual "XX Stopped working" Seems that the previous uppering voltage has degraded even more

I will call my wallet now


EDIT removing the dedicated video card and enabling the IGP, it artifacts like crazy in 3D apps

[micropage7]

so, have you checked every single ram?

[streetfighter 2]

Umm . . . Now I may be a simple hyper-chicken from a backwoods asteroid, but why not use memtest to test your RAM?

I mean, prime95 can detect bad RAM in much the same way that a thermometer can detect cancer. For all we know your CPU could be failing.

[Goodman]

As Neuromancer said PSU is a good possibility IMO

Is your inside PC look dusty , may need to clean it out? (PSU included)

Are you running your CPU overclock?

[Derek12]

Originally Posted by micropage7

so, have you checked every single ram?

I only have 1 stick of RAM installed and I don't have spare ones

Originally Posted by streetfighter 2

Umm . . . Now I may be a simple hyper-chicken from a backwoods asteroid, but why not use memtest to test your RAM?

I mean, prime95 can detect bad RAM in much the same way that a thermometer can detect cancer. For all we know your CPU could be failing.

I tried Memtest until the test 5 (I know very little time but I was afraid to leave it more) and none errors detected. But this wouldn't be the first time Memtest didn't detected errors and RAM was faulty.
Prime95 detects them instantly, as well as the continuous crashes in almost any app.
I am not sure the CPU is failing, because I previously upped RAM voltage to 1.9 and worked stable, and after that, now it's unstable even at 667.

Originally Posted by Goodman

As Neuromancer said PSU is a good possibility IMO

Is your inside PC look dusty , may need to clean it out? (PSU included)

Are you running your CPU overclock?

I will try to repair the cable for my previous PSU and I'll try (it's a tricky job anyway to solder 24 cables)

I blew the dust (despite the insides were pretty clean anyway) with canned air but the problem continues. No I didn't OCed the CPU but I tried underclocking it but it was unstable anyway.
And even the IGP generates artifacts (surely because uses RAM) in Atitool and Aero. All is fine using the dedicated card.

----

I will underclock it to DDR2-400 and see if it works fine.

Thanks to everyone


EDIT: I am getting even problems in POST, sometimes BIOS enters in "safe mode" and I have to press F1 to continue.

Once, a message saying "Bootmgr image is corrupt. The system cannot boot" appeared, but after restarting booted fine.

[Goodman]

Originally Posted by Derek12

I blew the dust (despite the insides were pretty clean anyway) with canned air but the problem continues. No I didn't OCed the CPU but I tried underclocking it but it was unstable anyway.
And even the IGP generates artifacts (surely because uses RAM) in Atitool and Aero. All is fine using the dedicated card.

----

I will underclock it to DDR2-400 and see if it works fine.

Thanks to everyone


EDIT: I am getting even problems in POST, sometimes BIOS enters in "safe mode" and I have to press F1 to continue.

Once, a message saying "Bootmgr image is corrupt. The system cannot boot" appeared, but after restarting booted fine.

Are you really using AtiTool?
it is to old use furmark instead for testing your GPU --> http://www.ozone3d.net/benchmarks/fur/

Still your PSU bug me lol!
Try another one if you can & a good one if all possible?

[Derek12]

Originally Posted by Goodman

Are you really using AtiTool?
it is to old use furmark instead for testing your GPU --> http://www.ozone3d.net/benchmarks/fur/

Still your PSU bug me lol!
Try another one if you can & a good one if all possible?

Yeah Atitool scanner is in my experience the far more sensitive piece of software despite being old (after you configure the sensitivity slider) it detects artifacts even when OCCT or atitraytools don't (and they aren't false positives ).

Anyway I used Furmark with the IGP and didn't noticing anything but in games and atitool artifacts appear because of the RAM I think.

Yeah I understand you this generic one is provisional I will shortly repair my good one and I will try it! But I think the problem is in the RAM itself is dying

Thanks

[Goodman]

Originally Posted by Derek12

Yeah Atitool scanner is in my experience the far more sensitive piece of software despite being old (after you configure the sensitivity slider) it detects artifacts even when OCCT or atitraytools don't (and they aren't false positives ).

Anyway I used Furmark with the IGP and didn't noticing anything but in games and atitool artifacts appear because of the RAM I think.

If you cant detect artifacts using Furmark (at least 10 min. running) then your IGP or GC are good
Don't trust what you see or not see with Atitool it is to old now last official version was in December 2006 & last Beta mid 2007...

My advice to you is STOP using Atitool , use Furmark to benchs or test your GC
As for GC overclocking software you can use MSI Afterburner or Trixx from Sapphire so please do remove/uninstall Atitool & forget about it...

I am pretty sure your ram is just fine , all i see here is...
1- Use of a bad/old software (Atitool)
2- PSU (Real possibility)
3- Voltage?

EDIT: You may as well try new version of Prime95 26.6 -->http://www.mersenne.org/freesoft/

[JrRacinFan]

Find another CPU to test with before calling anything else bad.

[Derek12]

Originally Posted by Goodman

If you cant detect artifacts using Furmark (at least 10 min. running) then your IGP or GC are good
Don't trust what you see or not see with Atitool it is to old now last official version was in December 2006 & last Beta mid 2007...

My advice to you is STOP using Atitool , use Furmark to benchs or test your GC
As for GC overclocking software you can use MSI Afterburner or Trixx from Sapphire so please do remove/uninstall Atitool & forget about it...

I am pretty sure your ram is just fine , all i see here is...
1- Use of a bad/old software (Atitool)
2- PSU (Real possibility)
3- Voltage?

EDIT: You may as well try new version of Prime95 26.6 -->http://www.mersenne.org/freesoft/

Many thanks for your advice and info I always used Atitool due to sensitivity but I will change to furmark/OCCT now

I measured PSU voltages and they are fine, in every moment. But my good one is almost repaired so maybe tomorrow I can test it, also I will borrow a known good DDR2-800 memory so the culprit should be know tomorrow maybe. I will update Prime95 too

Originally Posted by JrRacinFan

Find another CPU to test with before calling anything else bad.

Nope the only component I blamed along this thread is the RAM. I think motherboard and CPU are discarded, only remains RAM and PSU.
I can't get any AM2 CPU here, but I will borrow another DDR2-800 RAM I will comment the results.

[JrRacinFan]

Originally Posted by Derek12

Nope the only component I blamed along this thread is the RAM. I think motherboard and CPU are discarded, only remains RAM and PSU.
I can't get any AM2 CPU here, but I will borrow another DDR2-800 RAM I will comment the results.

Not necessarily. The extra voltage may be stabilizing the degrading IMC.

[OneMoar]

you need to prepair your wallet to buy a new pc ... cough ...
you are overclocking cheap ram to over 40% its rated speed
degradation is to be expected

[Derek12]

Originally Posted by OneMoar

you need to prepair your wallet to buy a new pc ... cough ...
you are overclocking cheap ram to over 40% its rated speed
degradation is to be expected

Buying a new PC.. Until the CPU and mobo dies, for now it works fine aside from this issue.

I only OCed to do benchmarks and after that I would let it at stock.

Originally Posted by JrRacinFan

Not necessarily. The extra voltage may be stabilizing the degrading IMC.

Today I will get the RAM, I will check if it's the CPU then I will be SOL.