Overclock.net › Forums › Intel › Intel - General › Asus Rampage Extreme IV + 3930k Problems
New Posts  All Forums:Forum Nav:

Asus Rampage Extreme IV + 3930k Problems

post #1 of 7
Thread Starter 
I have some issue that I am trying to track down and I wanted to provide some information and get some other opinions.

I am building/built a new rig using an Intel 3930k and a Asus Rampage Extreme IV and it is really unstable. I get memory management errors all the time and I can not run LinX, IBT or p95 for 5 minutes without crashing and getting said memory management errors. Otherwise the system runs semi stable during general use and only crashes a few times a day. In about 80% of the cases when I open the crash dumps It says the most likely culprit is memory corruption, further details show it happens during a read operation every time. The other 20% are random but usually lead back to a program could not access a requested part of mem etc. along with page fault in non paged area.

I have been through 3 separate memory kits, 2 os installs and removing all unnecessary hardware and a lot of trouble shooting and am beginning to believe it's the motherboard or CPU is defective. I will provide more information and thoughts on this below.

It's custom watercooled dual rad setup and I am only cooling the motherboard and cpu for now. My motherboard temps top out around 40C - 45C and my processor tops out around 60C under p95 etc. This is with the cpu clocked at 4.5ghz @ 1.4v max (0.080 offset), 1.2/ vtt/vcca. Medium - High LLC with around 120-130% CC. Nothing crazy mind you but I have tried a wide range testing incrementally.

I have tried with hyperthreading on and off and it seems to be more stable with hyperthreading off.

Running small FFT's in p95 is fine, no crashes. I have even been able to get regular blend to run for up to 6 hours with higher dimm voltage and ridiculously relaxed timings but it doesn't really use a lot of memory. Anything custom using more memory crashes within seconds to minutes. If I use over 3500mb in LinX same deal.

I have also done all the trouble shooting at stock / optimized default etc with small tweaking. Ram at stock timings using xmp, manual. Tweaking vdimm up a bit etc. loosened timings, tightened timings. Pretty much every thing that has been posted on any forum that google could find that wasn't just crazy.

Originally I thought it might be the mem because I bought a gskill 2400 1.65v kit. It would not run at the rated speeds and even at slower speeds I would still get the memory management errors. I read the 3930k like lower voltage mem so I went out a bought a couple 8gb kingston hyperx blu 1600 1.5v sticks. They work better but still had to loosen the timings and tweak vdimm up to about 1.55 to get them semi stable. Loosening tertiary timing seems to have the best effect but at 6 - 15 it's way below what should be needed for 2 dimms and the errors still persist. I went out and bought a 3rd kit that was rated for more speed at 1.5v thinking that in a worst case scenario I have the crappiest IMC on the planet and I could run the ram at normal timings at a slower speed, 1600 per intel spec. It is a Patriot Viper III 2133, 1.5v designed for intel processors. It was the recommend kit on their website and I have always had pretty good luck with patriot memory.

Still the same issue persists with the exact same symptoms using 3 different kits in every configuration possible. I have been going through this for almost 3 weeks.

On a side note I did check the ram using memtest86+ in multiple configurations including single sticks through all of this and never got one error but never ran it more than 2-3 runs.

So I decided to try things out using just 1 x 8gb stick. Same issues. So I decided to check the individual dimm slots and got really varying results. On slot A1 the ram can load windows and run LinX with 3500mb and crashes after 5 minutes, slot B1 loads windows and runs LinX for 10 minutes, slot D1 can't load windows, slot C1 loads windows but crashes before I can really do anything.

I am really at the point of pulling my hair out now. lol.

So, I completely disassembled everything to remount my blocks, check for any bent pins, look for any visible damage, cleaned everything and put everything back together. No problems were found. Although I did get a better spread on my TIM this time using a different method which dropped my CPU temps a few degree's. wink.gif

So if anyone has any further questions or advice. Please let me know. lol

I am not sure if the crazy differences in RAM working or not working in different slots points to a motherboard issue or if I could still be a CPU issue? What I do know is that it's definitely not a RAM issue. I don't exactly have 500$ motherboards or processors lying around all over the place and need to figure out whether it's the CPU that needs replacement or the motherboard.

Any advice is greatly appreciated.
post #2 of 7
Thread Starter 
Also, one other reason I though it might be motherboard is some weird glitch in the gfx of the uefi bios, one of the tab headers was corrupted. I updated the bios and it seems to fix the glitch. Not sure if that could be an indicator of some other issues with the board.

If I hovered over the tab to highlight it it would show the correct focused image but when I moved the mouse back it would flip back to the corrupted image.


Edited by zer0patches - 3/31/13 at 11:01am
post #3 of 7
If it does it at stock/optimized defaults, rma it.
Kill 2 birds with 1 stone, mobo and cpu at the same time.

But have you tried different bios revisions?
Inspected the skt for bent pins?
Even loosened the wb a couple of turns?
Mighty-iTX
(12 items)
 
One foot in..
(17 items)
 
mATX
(12 items)
 
CPUMotherboardGraphicsRAM
i7-4770K M7-Impact Strix GTX980 Samsung 
Hard DriveCoolingOSMonitor
Crucial M4 256 Glacer 240L W8.1 K272HUL 
KeyboardPowerCaseMouse
Saitek Seasonic X-650 Prodigy MX518 
CPUMotherboardGraphicsGraphics
4770K/45/43 Maximus VI Extreme R290X R290X 
RAMHard DriveHard DriveOptical Drive
Samsung Crucial M4 64GB RAIDR LSSWM 
CoolingOSMonitorKeyboard
Cooler Master Glacer 7 Ultimate 64 SyncMaster Eclipse 
PowerCaseMouseMouse Pad
XFX 1050 BE 600T Silver MX518 Desk 
CPUMotherboardGraphicsRAM
i5-4330 Z87 Gryphon GTX690 G.Skill RipJawsX 
Hard DriveCoolingOSMonitor
WD Blacks Hyper 212 W8.1 ACER 
KeyboardPowerCaseMouse
Logitech Ultra X3 1000w Corsair 230T Orange Razer 
  hide details  
Reply
Mighty-iTX
(12 items)
 
One foot in..
(17 items)
 
mATX
(12 items)
 
CPUMotherboardGraphicsRAM
i7-4770K M7-Impact Strix GTX980 Samsung 
Hard DriveCoolingOSMonitor
Crucial M4 256 Glacer 240L W8.1 K272HUL 
KeyboardPowerCaseMouse
Saitek Seasonic X-650 Prodigy MX518 
CPUMotherboardGraphicsGraphics
4770K/45/43 Maximus VI Extreme R290X R290X 
RAMHard DriveHard DriveOptical Drive
Samsung Crucial M4 64GB RAIDR LSSWM 
CoolingOSMonitorKeyboard
Cooler Master Glacer 7 Ultimate 64 SyncMaster Eclipse 
PowerCaseMouseMouse Pad
XFX 1050 BE 600T Silver MX518 Desk 
CPUMotherboardGraphicsRAM
i5-4330 Z87 Gryphon GTX690 G.Skill RipJawsX 
Hard DriveCoolingOSMonitor
WD Blacks Hyper 212 W8.1 ACER 
KeyboardPowerCaseMouse
Logitech Ultra X3 1000w Corsair 230T Orange Razer 
  hide details  
Reply
post #4 of 7
I'm pretty sure the problem is that you haven't tried increasing the memory controller voltage (VVT). Probably the culprit. thumb.gif
post #5 of 7
Thread Starter 
Quote:
Originally Posted by grunion View Post

If it does it at stock/optimized defaults, rma it.
Kill 2 birds with 1 stone, mobo and cpu at the same time.

But have you tried different bios revisions?
Inspected the skt for bent pins?
Even loosened the wb a couple of turns?

The problem is this is my primary PC and I am working on a big project. Right now I can't afford to RMA both, it's just passed 30 days now so I would need to go to the manufacturer and no telling how long that takes. I was hoping for an upfront exchange where I just give them my cc # and they give me 14 days to send the old unit in.

Yes, I have tried 3 bios versions. The original that came on the board, the one from December 2012 and also the latest one from march 2013. I have also tried both bios slots since the MB comes with 2.

Yes, I disassembled everything, cleaned, checked for bent pins or any other damage. I checked the bottom of the processor and it looks likes the pins are lining up well. Everything was remounted and I did not find any view-able damage.

Yes, I loosened several turns until the temps went up about 5C to make sure it was loose enough. It did not help, just increased my temps.
post #6 of 7
Thread Starter 
Quote:
Originally Posted by BiG StroOnZ View Post

I'm pretty sure the problem is that you haven't tried increasing the memory controller voltage (VVT). Probably the culprit. thumb.gif

I thought the memory controller on the 3930k was handled by the system agent, vccsa in my bios. I have tried up to 1.325v briefly . Intel max spec is 1.4 volt but I am C2 stepping and I have read that leaving vccsa over 1.2 long term will degrade the processor more quickly. I have kept vtt and vccsa the same for stability reasons, normally 1.200v which is the max recommended and should easily support 2133mhz ram.

I have even tried the ram at stock 1600mhz profile with the voltages upped a bit and it's the same problem no matter which kit or how many sticks I have used. I know out of the 14 sticks I have tried in the machine they are not all bad. =p

Thanks for the reply guys.

I am a bit miffed, I have been building pc's for 20 years since the 386/486 days. I used to be microsoft certified and did field service repairing computers, I also worked for a system builder so I have literally built / fixed hundreds of computers and have a very wide range of experience. I built myself a new 4u server last year and went with a intel 2600k, I was pretty impressed with the build, it performed better than my amd 1100t hexcore. I have always been a fan of amd from a price / performance standpoint but decided to go intel with my new personal machine since I wanted to go water cooled and I like the overclocking headroom on the newer intels.

I am about at the point where I am just gonna go buy another lga 2011 processor, probably the cheaper quad core to see if it's my processor or the motherboard. It's a pain to pull the system apart since I am actually using it and just dealing with the crashes. At least if it is the processor I will have one to use while I rma the other one and will be able to stay up and running.
Edited by zer0patches - 3/31/13 at 4:41pm
post #7 of 7
Thread Starter 
Well, I bit the bullet and went out and bought an LGA2011 i7 3820 in order to figure out if it's my processor or motherboard.

Looks like I have a bad 3930k.

No more memory crashes and can run large mem in LinX etc now. My memory's XMP profile even works at 2133 out of the box. Time to RMA my 3930k with Intel. mad.gif

I'm just thankful it's not my motherboard. lol
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel - General
Overclock.net › Forums › Intel › Intel - General › Asus Rampage Extreme IV + 3930k Problems