M5A97 LE R2.0 MOBO with GeForce 680x and RipJaws (2) 8GB RAM experiencing random freezes and lock ups

[SonnyBlue]

Hey guys, so it's been a long battle and I finally want to come to the experts and ask. My rig will run fine for around 2-3 hours, and than randomly, it will lock up, and force me to hard reset. My current configuration is a M5A97 LE R2.0 MOBO with GeForce 680x and RipJaws (2) 8GB RAM. I am just lost as what to look for after updating both my BiOS and graphic card drivers repeatedly. I'm aware it could be the RAM, is there a way to check that without just buying all new sticks? Hopefully y'all can help me out. Thanks

 

[Calvin7]

GeForce GTX 680
Minimum System Power Requirement: 550 watts

RipJaws (2) 8GB RAM
Do you have part number or link to the memory?

CPU model ?

 

[SonnyBlue]

GeForce GTX 680
Minimum System Power Requirement: 550 watts

RipJaws (2) 8GB RAM
Do you have part number or link to the memory?

CPU model ?

RipJaws are part number: G.SKILL Ripjaws X Series 8GB (2 x 4GB) 240-Pin DDR3 SDRAM DDR3 1600 (PC3 12800) Desktop Memory Model F3-12800CL9D-8GBXL

And the CPU is: AMD FX(tm)-8350 Eight Core.

 

[MeanMachine41]

Your G.SKILL Ripjaws X Series 8GB (2 x 4GB) are not a mismatch, however that does not rule out a failing module. You can also check the MB QVL to check they have been tested.
Are the 16GB RAM modules composed of two 8GB kits.? If so this can be an issue as its never a good idea to mix kits,even of the same spec.
RAM is binned in kits at the manufacturers factory to ensure latency compliance due to minor imperfections.
Run memtest86+ from a USB stick to rule out the possibility of DIMM failure.
If all DIMMs pass 3 passes on all modules then Stress testing should find the culprit.

For Stress Testing download HWMonitor and AIDA64 and put them side by side on your desktop. I'll guide you where to go from there.

 

[Calvin7]

(2) 8GB

Combining multiple kits can cause the PC to be unable to boot or to be unable to operate at rated specifications.

G.Skill does not guarantee such combinations will work. Trying to run at lower memory clock speed may get them to run.

A single kit does have a lifetime guarantee.

Also, MemTest86+ has not been supported since Sept. 2013. It does not support UEFI.
You might want to use MemTest86.
http://www.memtest86.com/

 

[SonnyBlue]

Your G.SKILL Ripjaws X Series 8GB (2 x 4GB) are not a mismatch, however that does not rule out a failing module. You can also check the MB QVL to check they have been tested.
Are the 16GB RAM modules composed of two 8GB kits.? If so this can be an issue as its never a good idea to mix kits,even of the same spec.
RAM is binned in kits at the manufacturers factory to ensure latency compliance due to minor imperfections.
Run memtest86+ from a USB stick to rule out the possibility of DIMM failure.
If all DIMMs pass 3 passes on all modules then Stress testing should find the culprit.

For Stress Testing download HWMonitor and AIDA64 and put them side by side on your desktop. I'll guide you where to go from there.

I have both programs now, if you have time to help me that would be great.

 

[MeanMachine41]

Your G.SKILL Ripjaws X Series 8GB (2 x 4GB) are not a mismatch, however that does not rule out a failing module. You can also check the MB QVL to check they have been tested.
Are the 16GB RAM modules composed of two 8GB kits.? If so this can be an issue as its never a good idea to mix kits,even of the same spec.
RAM is binned in kits at the manufacturers factory to ensure latency compliance due to minor imperfections.
Run memtest86+ from a USB stick to rule out the possibility of DIMM failure.
If all DIMMs pass 3 passes on all modules then Stress testing should find the culprit.

For Stress Testing download HWMonitor and AIDA64 and put them side by side on your desktop. I'll guide you where to go from there.

I have both programs now, if you have time to help me that would be great.

OK go into the Tools section of AIDA64 and select to run the stress test.
Tick the boxes for CPU, FPU and Cache. Run the test for 10mins and stop the test if temperatures reach 80C.
Later you can also run tests for other sub systems if you wish. Run test on RAM separately.
Constantly monitor HWMonitor for system temperatures and do not leave unattended.
At the 10min mark, take print screen shots and upload pics to your favorite file host.
You can use IMIGUR as a host, upload the files then go to images. Here you can obtain the BBurl to link here for analysis.
If you can't get all info, page down or change menu screens and take more shots. Select those pics you wish to have analysed.

I would want to see all rail voltages during the test with 100% utilization on all cores.

Try with just one kit of matching RAM initially.

 

[SonnyBlue]

Your G.SKILL Ripjaws X Series 8GB (2 x 4GB) are not a mismatch, however that does not rule out a failing module. You can also check the MB QVL to check they have been tested.
Are the 16GB RAM modules composed of two 8GB kits.? If so this can be an issue as its never a good idea to mix kits,even of the same spec.
RAM is binned in kits at the manufacturers factory to ensure latency compliance due to minor imperfections.
Run memtest86+ from a USB stick to rule out the possibility of DIMM failure.
If all DIMMs pass 3 passes on all modules then Stress testing should find the culprit.

For Stress Testing download HWMonitor and AIDA64 and put them side by side on your desktop. I'll guide you where to go from there.

I have both programs now, if you have time to help me that would be great.

OK go into the Tools section of AIDA64 and select to run the stress test.
Tick the boxes for CPU, FPU and Cache. Run the test for 10mins and stop the test if temperatures reach 80C.
Later you can also run tests for other sub systems if you wish. Run test on RAM separately.
Constantly monitor HWMonitor for system temperatures and do not leave unattended.
At the 10min mark, take print screen shots and upload pics to your favorite file host.
You can use IMIGUR as a host, upload the files then go to images. Here you can obtain the BBurl to link here for analysis.
If you can't get all info, page down or change menu screens and take more shots. Select those pics you wish to have analysed.

I would want to see all rail voltages during the test with 100% utilization on all cores.

Try with just one kit of matching RAM initially.

Here is the link to the results: http://imgur.com/gallery/4q1Tf

Again, thanks for all of your help.

 

[MeanMachine41]

Yea, best to not make your Images public as many don't know what they are looking at and can't analyse the results. The person commenting on your 3.3V rail is quite correct however, your PSU is not supplying an efficient 3.3V supply. Acceptable tolerances are +/- 5% on all rails. Hardware effected by the 3.3V rail are the chipsets, RAM (DIMM) and some PCI/AGP cards.

Your temperatures are fine at 4.2GHz.

If you can, Swap out the PSU with a known working unit of the same Wattage or higher or have your unit tested under load. That would eliminated the PSU as the problem. You can't do it yourself without sophisticated equipment. You can only test with a mutimeter however its when under load what counts.

Now run a test on the GPU for 30mins and link images here.

Next would be the DIMMs, test as single kit and both kits for 1hr. As your getting a low 3.3V it may affect the results.

 

[SonnyBlue]

Yea, best to not make your Images public as many don't know what they are looking at and can't analyse the results. The person commenting on your 3.3V rail is quite correct however, your PSU is not supplying an efficient 3.3V supply. Acceptable tolerances are +/- 5% on all rails. Hardware effected by the 3.3V rail are the chipsets, RAM (DIMM) and some PCI/AGP cards.

Your temperatures are fine at 4.2GHz.

If you can, Swap out the PSU with a known working unit of the same Wattage or higher or have your unit tested under load. That would eliminated the PSU as the problem. You can't do it yourself without sophisticated equipment. You can only test with a mutimeter however its when under load what counts.

Now run a test on the GPU for 30mins and link images here.

Next would be the DIMMs, test as single kit and both kits for 1hr. As your getting a low 3.3V it may affect the results.

http://imgur.com/p1bPci1
http://imgur.com/TwOaxj3

So, I couldn't even run the stress test on the GPU for more than 7 minutes, as I was having the lock up/reset issue. Looks like it's time to upgrade my graphics card?

 

[MeanMachine41]

The test is inconclusive for your GPU as your pics only show graphs detailing CPU and temperature readings.
You should have tested for 30mins and take shots with the results of Voltage, Powers, Clocks, and Statistics. The card may be the problem but that has not been ruled out yet.
The GTX 680 is still a viable card and there are other GPU specific tests we can do before you upgrade.

I would also like to see all rail voltages 12V, 5V, 3.3V in HWMonitor during the test to see if your PSU is doing its job as i'm still suspecting the 3.3V rail.