Go Back   OC3D Forums > [OC3D] Processors & Platforms > Memory
Reply
 
Thread Tools Display Modes
 
  #1  
Old 11-05-13, 09:14 PM
Vicey Vicey is offline
OC3D Crew
 
Join Date: Apr 2012
Location: UK
Posts: 747
Lets talk: RAM degradation over time

I've been building my own computers for a decade now and one thing I've noticed throughout that time from DDR1 to DDR3, over 150GB+ of memory that I've owned is that over time RAM degrades.

Now obviously all things degrade but I've noticed that RAM degrades very quickly whilst being solid state. Much faster than motherboards, CPU's and GPU's do (Not including fan life obviously, just the silicon).

On more than one build I've done I have had to tweak voltages ever higher over the life of the builds for memory. There is always some point usually after 12 months of ownership where by the system randomly is no longer stable. I never overclock my RAM if anything I underclock it more often than not to get higher CPU performance.

And it occurred to me that people don't really discuss this or at-least I've not seen any threads on it so I'd thought I would start one so we can discuss it. The reason I bring it up at all is because it's happened to me (again) with my current RIVE build in my signature. The 32GB of Mushkin Memory is almost 14 months old now and today I just couldn't get it stable at the recommended specs which it had operated at fine since day one.

Recommended:
1866MHz / 9-9-9-27 / 1.500v

What I have to do now to stay stable:
1866MHz / 9-9-9-27 / 1.550v

Now when I say stable I mean the system would crash booting up, shutting down (if it did manage to boot up), opening my browser crashed the system and Prime95 would error out within 1 second with rounding errors detected on every thread. Changing the voltage to 1.550v however a 0.050v increase makes the system completely Prime95 stable for over 2 hours so far. I could not get it to run more than a couple seconds at stock.

So have any of you experienced the same thing? And what were your conditions did you have your RAM overclocked or at stock or underclocked?

__________________
Intel Core i9 7900X @ 4.7GHz [Delidded & Lapped] @ 1.23v Asus Rampage VI Extreme G.Skill Trident RGB 32GB 3600MHz CL17 Samsung 1TB 970 Pro NVIDIA GTX 1080 Ti FE @ 2GHz Asus Swift PG279Q Custom Water Loop
Reply With Quote
  #2  
Old 12-05-13, 01:07 AM
alienware alienware is offline
Banned
 
Join Date: Jun 2010
Location: On Mars as an anthropologist
Posts: 0
Send a message via MSN to alienware
Nope never had an issue with it tbh.

There's a chance that your memory just isn't happy with 1.5v or the board you are using doesn't have a stable memory voltage.
Reply With Quote
  #3  
Old 12-05-13, 01:33 AM
Vicey Vicey is offline
OC3D Crew
 
Join Date: Apr 2012
Location: UK
Posts: 747
Quote:
Originally Posted by AlienALX View Post
Nope never had an issue with it tbh.

There's a chance that your memory just isn't happy with 1.5v or the board you are using doesn't have a stable memory voltage.
Like I said in my post, I've run this rig for almost 14 months at 1.5v perfectly. It is just today that the RAM has degraded to a point where I need to give it 1.55v to keep it stable.

My boards voltages are perfect, it is definitely the RAM. And this isn't the first time I've seen this. I've probably had this happen 4 or 5 times over the past decade and it usually happens with the high end memory the ones that cost a lot and are being cherry picked for their speeds at the factory.

I've had this happen on OCZ, Crucial, Corsair and now Mushkin across all different boards from AMD nForce 4 boards to NVIDIA nForce 6 boards to Intel X58 and X79 boards.

I think in some ways the high end memory is very much in the same ballpark as some of the cheap SSD's on reliability.
__________________
Intel Core i9 7900X @ 4.7GHz [Delidded & Lapped] @ 1.23v Asus Rampage VI Extreme G.Skill Trident RGB 32GB 3600MHz CL17 Samsung 1TB 970 Pro NVIDIA GTX 1080 Ti FE @ 2GHz Asus Swift PG279Q Custom Water Loop
Reply With Quote
  #4  
Old 12-05-13, 05:15 PM
CloneageDK CloneageDK is offline
Newbie
 
Join Date: Sep 2011
Location: Denmark
Posts: 48
I have 6 sticks of Kingston HyperX T1 in my X58 system, advertised at 1600MHz with 1.65v. They could do that without problems in the beginning, but I find myself struggling to get them to do that now. Either the PC won't POST, Windows fails to boot up completely or I get BSOD's.
I haven't played with voltages too much though, I just run them at 1333MHz instead and it's stable. Mostly

I have 3 sticks of Corsair memory advertised at 1333MHz that I used before the Kingston RAM and while being considerably older (and having been used a lot more), they still do that today.

I no longer have spare parts lying around so I can't really play too much with it, but the Kingston RAM started to get issues on a former Asus X58 motherboard I had and they show the same problems now on my current Gigabyte X58 motherboard. Again, the Corsair RAM are just fine.

With all of that being said, I haven't had problems with it before or after. Only with these Kingston RAM ^^
Reply With Quote
  #5  
Old 12-05-13, 08:00 PM
alienware alienware is offline
Banned
 
Join Date: Jun 2010
Location: On Mars as an anthropologist
Posts: 0
Send a message via MSN to alienware
Quote:
Originally Posted by Vicey View Post
Like I said in my post, I've run this rig for almost 14 months at 1.5v perfectly. It is just today that the RAM has degraded to a point where I need to give it 1.55v to keep it stable.

My boards voltages are perfect, it is definitely the RAM. And this isn't the first time I've seen this. I've probably had this happen 4 or 5 times over the past decade and it usually happens with the high end memory the ones that cost a lot and are being cherry picked for their speeds at the factory.

I've had this happen on OCZ, Crucial, Corsair and now Mushkin across all different boards from AMD nForce 4 boards to NVIDIA nForce 6 boards to Intel X58 and X79 boards.

I think in some ways the high end memory is very much in the same ballpark as some of the cheap SSD's on reliability.
And you're sure it's the ram? have you ever tried it in a different motherboard?

As I say I've never had it happen to me but I have had board's capacitors degrade over time and slacken off on their power delivery.
Reply With Quote
  #6  
Old 12-05-13, 09:09 PM
NeverBackDown NeverBackDown is offline
AMD Enthusiast
 
Join Date: Dec 2012
Location: With the Asguardians of the Galaxy
Posts: 16,211
You sure it is not your Cpus memory controller starting to go out?
__________________
I am Iron Man.
Reply With Quote
  #7  
Old 13-05-13, 02:59 AM
Bartacus Bartacus is offline
OC3D Elite
 
Join Date: Apr 2013
Location: Ottawa, ON, Canada
Posts: 2,307
I keep my rigs for 2-3 years, and sell them when I upgrade, and I've never seen this.
__________________
Guts: Ryzen 2700X / Asus CH7 Hero / 32GB TridentZ RGB 3200CL14 / Zotac GTX 1080Ti x2 / EVGA 1000W PSU / Case Labs M8
Storage: ADATA SX8200 1TB NVME / 6TB RAID0 HDD array / 4TB RAID0 SSD array
Water Cooling: Watercool HeatKiller IV blocks on CPU & GPUs / 4 360 rads / Barrow box reservoir / dual D5 pumps

Reply With Quote
Reply

Tags
degradation, ram, voltage

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump










All times are GMT. The time now is 06:53 AM.
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.