Asrock 970 Extreme4 AM3+ not stable at stock :( :(

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic

 DarkStar-82 wrote:

Hi there, before I pull this thing out of my system 
and set fire to it I thought I would come on and ask people who may be 
more knowledgeable about this particular piece of... hardware... than I 
am.

The basic story is this, About 6 months ago I bought this 
board. I planned to buy a Gcard a few months after and game away. Ive 
experienced a system halt, not blue-screen, a complete hang/freeze with 
loud screeching, at least twice a week, but I had put this down to bad 
drivers ect ect.

Well 2 months ago I decided to get ready for the
 game Ive been waiting for, StarCitizen. I picked up an R9 290 and 
retired my HD5870. I upgraded my CPU from an FX6300 to an FX8320. I 
added another 8 gig of matched branded Gskill 1600 Ram. 

And I can play precisely nothing.  
 I have 3 titles, I play them on my APU laptop without much issue, sure 
the games crash at times but the entire system doesnt halt, nor should 
it unless there is a serious problem somewhere. 

So heres my spec:
Asrock Extreme4 970 board @ stock settings - bios 2.80
FX8520 8core 3.5Ghz. Watercooled. No overclocking! I want silence, not speed.
16gig (4x4gig) Gskill Ripjaw DDR3 1600Mhz
Sapphire R9 290 Graphics card - Stock speeds with factory fitted Xsilence cooler.
1xSamsung SSD
1xconventional Seagate HDD
5x 120mm fans
2X Antec Gamer 500watt PSU. One dedicated to graphics card. One to system.

My Spares Ive used also in testeing:
FX6300
HD5870
380watt Seasonic PSU

Software used in testing:
Passmark
Prime 95
Intel burn in test
cpu-burn-in

Heres what Ive tried,
 (after each change I have ran http://cpuburnin.com on a bootable CD to 
eliminate any possible issues from drivers or Windows, and each time it 
has failed with a system lock. i have been able to replicate the lock 
using other bootable CPU testing tools in a similar time frame hence I 
do not believe there is any issue with the program.

Disconnected all hardware / wires apart from graphics and 1 ram chip.
Changed CPU.
Changed through all ram chips.
Changed graphics card.
Changed graphics card PCI-E socket location to secondary.

Bios changes attempted, both CPU`s and graphics cards tested in system, made no difference:
a)increased
 vcore from auto to 1.25 through to 1.4v testing with each incremental 
change. System was more unstable. Lower iterations reached.

b)lowered
 Northbridge and south bridge voltages+ frequencies, system ran CPUtest 
 iterations to much higher numbers this time but still failed within 10 
minutes.

c)increased nb Voltage by 0.25v in increments and tested. System was more unstable. Lower iterations reached.

d) I dragged a 20" 5 foot fan down two floors, pointed it two feet away, and blew a tornado at it.
no luck there either.

So basically Ive got a gaming Pc thats good for watching TV, most of the time.

Any help appreciated. Thanks for your time.

Going by the date and time of the first and second posts of the OP, he gave up after ~14 hours of not getting a response.

The comment in the OP's first post, "before I pull this thing out of my system 
and set fire to it", speaks for itself.

Once again, we have a budget AMD board being used with an FX-8300 series processor. 

Intel is hated by many PC enthusiasts for their "Z-only" chipset board rules for their high TDP processors and over clocking, which is said to be a "tax" on that hardware. There is more to it than meets those user's eyes, as in making sure the boards are up to the task of using any compatible processor. 

AMD's "anything goes", no rules is loved by their customers, but in reality AMD is NOT doing them a favor and not looking out for their best interest IMO. The irony IMO is Intel is the one saving some users from themselves.

The 970 Extreme4 board is not marketed as a gaming board.

Water cooling is mentioned by the OP. No details beyond that. Regardless, we likely have yet another case of an AIO CPU cooler that provides no air flow over the CPU VRMs. The VRM stage is a 4 + 1 design, with a small heatsink. Add an FX-8320, and we have the classic scenario for VRM over heating. That does not include the 970 chipset over heating, which provides the PCIe lanes for the video cards.

No mention of the PC case, I'm guessing given the dual PSUs being used, an open test bench of some kind? Regardless, aiming a 20" fan at the PC does not guarantee proper cooling to the CPU VRM heatsink.

No mention of CPU temperature, or if the VRM heatsink is hot to the touch. Given the testing using the "http://cpuburnin.com on a bootable CD", that seems to be a very intensive CPU stress test. The following is from the CPU Burn-in home page:

CPU Burn-in v1.01 by Michal Mienik is the ultimate stability testing tool for overclockers.  
    The program heats up any x86 CPU to the maximum possible operating temperature that is achievable
    by using ordinary software.  This allows the user to adjust the CPU speed up to the practical
    maximum while still being sure that stability is achieved even under the most stressful conditions.
    The program continuously monitors for erroneous calculations and errors ensuring the CPU does not
    generate errors during calculations performed under overclocking conditions.

    New! Version 1.01 adds second mode that stress tests CPU even further.

This program is not commonly used by the over clocking community, so very little is known about it. If it does heat stress a CPU as much as it claims to do, then over heating of the CPU and/or VRM stage is possible.

Component temperatures of the CPU, chipset, and VRM heat sink is critical information needed to begin diagnosing the OP's issue, but that data is completely missing.

If the OP wants to continue this discussion, we will be happy to do so.