Computer freezing(FX 9590 and ASUS Crosshair V Formula-Z)

[PhoenixWright]

Hello everyone!
I wanted to post this thread from my computer but i cant do it if my computer isn't working, and thats why im here asking for help.

So since almost 7 months i got a new gaming computer, but i've faced tons of problems and i solved all of them, but there is one problem that is very persistent and it won't get fixed no matter what i did.

The problem is freezes caused by an unknown source that i've been trying to find.
I've done a memtest and i got no errors so we could exclude the ram, i also did a stress test but the computer froze so i cant know if it was caused by the cpu.

I also tried to disable c1 and c6 state on the bios but it didnt fix the problem, i think i can exclude the cpu also because when i did the stress test, i had hwmonitor open to check if the cpu voltage would be stable when the computer freezes, which was stable (1.46-1.5 Vcore).

So now, i cant find what is causing the freeze and i also cant fix it, i tried many solutions all over the internet. Many people have the exact same conditions but non of their solutions fixes my computer.

The freezes always happen when im playing games for half an hour and more and sometimes it happens while the computer is running windows.
It also happens from the first second i start a blend test on prime95.

Computer specifications is on the signature.

 

[gregvibe]

I think you should check the temps on your cpu and your entire system first of all.

 

[lunyone]

Well my initial thought is that you might have a temperature problem, not sure if it is Mobo, GPU, RAM or whatever the issue. Then I thought that you might have a PSU issue and then noticed that you have a 1300w PSU. So here is some things that I'd consider trying, if not already tried them:

* Check CPU thermal paste on the CPU, maybe you have too much and it has flowed down onto some of the Mobo parts and could be causing issues.
* Pull your GPU out of the system and inspect the contact points on the GPU and also in the slot that you put the GPU into the Mobo for any debris or anything that looks a bit odd.
* Re-install the GPU and this time try and use 2 different PCI-e power connectors for the GPU (I think you have 6 x 6+2 PCI-e power connectors, so you would have 3 options to try initially).
* Double check that your Mobo has the latest BIOS version, I think it is Version 2201 found HERE.
* Verify that you have sufficient cooling of your parts in the case (i.e. case fans operational and that you can feel/see the fans moving on the GPU and Case).

 

[PhoenixWright]

Well my initial thought is that you might have a temperature problem, not sure if it is Mobo, GPU, RAM or whatever the issue. Then I thought that you might have a PSU issue and then noticed that you have a 1300w PSU. So here is some things that I'd consider trying, if not already tried them:

* Check CPU thermal paste on the CPU, maybe you have too much and it has flowed down onto some of the Mobo parts and could be causing issues.
* Pull your GPU out of the system and inspect the contact points on the GPU and also in the slot that you put the GPU into the Mobo for any debris or anything that looks a bit odd.
* Re-install the GPU and this time try and use 2 different PCI-e power connectors for the GPU (I think you have 6 x 6+2 PCI-e power connectors, so you would have 3 options to try initially).
* Double check that your Mobo has the latest BIOS version, I think it is Version 2201 found HERE.
* Verify that you have sufficient cooling of your parts in the case (i.e. case fans operational and that you can feel/see the fans moving on the GPU and Case).

The idle temperature is 25-35, and when i stress the CPU it reaches 50-65. I don't think this is a dangerous temperature.
I also updated the bios to 2201 which didn't fix the problem, I also had checked all fans and they are running fine and also everything is cool including the GPU. (40-50 at idle and 55-70)
I tried the GPU solution you mentioned, but it didn't quite change anything.

 

[lunyone]

Okay I found THIS thread that you might to peruse to see if any of the things that they have done will help you with your "similar" situation. Long story short is that the CPU likes to be at mostly stock settings, along with the RAM, and that you might be getting some voltage drop on the CPU after a period of time. Check out that thread and see if that helps.

 

[PhoenixWright]

Okay I found THIS thread that you might to peruse to see if any of the things that they have done will help you with your "similar" situation. Long story short is that the CPU likes to be at mostly stock settings, along with the RAM, and that you might be getting some voltage drop on the CPU after a period of time. Check out that thread and see if that helps.

I actually have applied MOST of the solutions they recommended on that thread and it didn't fix anything, I tried them yesterday. but i wasn't able to try some of them because i didn't have the required tools or time.

 

[PhoenixWright]

Ok, so i did a stress test again and this time it lasted a little bit longer.
it lasted longer but one of the workers stopped working *1 errors, 0 warnings*
this is the error(included with the errors i got yesterday).
[Sun Nov 29 03:11:04 2015]
FATAL ERROR: Resulting sum was 5.16438017565826e+019, expected: 1.401953296443072e+016
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Resulting sum was 3.121754748831662e+016, expected: 7.698393639740254e+016
Hardware failure detected, consult stress.txt file.
[Mon Nov 30 07:36:36 2015]
FATAL ERROR: Resulting sum was 3.867419824825041e+018, expected: 3.576708232458507e+016
Hardware failure detected, consult stress.txt file.

the monday text is an error on worker #4
the one before it is an error on worker #3 and #4

 

[Darkbreeze]

Are you using liquid cooling, and if so, what is your case model and what does your case fan configuration look like. The #1, above all other concerns, problem, with the FX-9590 is inadequate power supply use. The next one is people using liquid cooling but very little case airflow and VRM thermal throttling even with very good motherboards.

What is your power supply model number?

How many case fans, radiator fans, etc., do you have installed and in what configuration? Just because other temps (CPU, GPU) don't reflect a temperature issues doesn't mean the VRMs are getting enough airflow over them and with liquid cooling you don't get the residual airflow over them like you would with an air cooler. Try removing the side panel temporarily and see if that helps things at all. If you remove the side panel and place a small electric fan pointed at the motherboard and it resolves or alleviates some of the issue, or all, you know you don't have enough airflow through the case.

This is extremely common with the FX-9590. I've dealt with like ten of these threads just this week.

 

[PhoenixWright]

Are you using liquid cooling, and if so, what is your case model and what does your case fan configuration look like. The #1, above all other concerns, problem, with the FX-9590 is inadequate power supply use. The next one is people using liquid cooling but very little case airflow and VRM thermal throttling even with very good motherboards.

What is your power supply model number?

How many case fans, radiator fans, etc., do you have installed and in what configuration? Just because other temps (CPU, GPU) don't reflect a temperature issues doesn't mean the VRMs are getting enough airflow over them and with liquid cooling you don't get the residual airflow over them like you would with an air cooler. Try removing the side panel temporarily and see if that helps things at all. If you remove the side panel and place a small electric fan pointed at the motherboard and it resolves or alleviates some of the issue, or all, you know you don't have enough airflow through the case.

This is extremely common with the FX-9590. I've dealt with like ten of these threads just this week.

My case is a NZXT H440 Red Here,
and I'm using water cooling thermaltake water 3.0.

I have 3 fans in the front of the case and 1 CPU cooler in the rear. so 4 at an overall.
my power supply is an EVGA Supernova G2 1300 watts.
I don't have an electric fan but the case side panel is open and the room is pretty much very cold(I'm using an air conditioner because i live in UAE).

P.S. Darkbreeze, You have helped me before on the thread before this, HERE, So this is the second time you are helping with the freezing problems 😛

 

[Darkbreeze]

I don't mean to seem dense, but I'm not sure what you mean by "1 CPU cooler in the rear". You are either using water cooling or air, not both. Do you mean 1 case fan in the rear?

I see that I helped you before, along with Damric, but it's been a while. Let's start fresh shall we?

I understand that people buy these processors because of the advertised specs, but the fact is that the additional clock speed doesn't really help that much and the voltage necessary to achieve those speeds tends to create too much heat and stress on the motherboard VRMs, memory controller and northbridge.

My recommendation would be to cap the CPU at about 4.5Ghz using the BIOS settings.

Also, if you only have a single exhaust fan installed, and your radiator is top mounted as an intake, you are definitely not getting enough heat OUT of the case fast enough to keep temps down or create the necessary draft across the top of the motherboard to help cool the VRMs since you're using water for the CPU and are not getting any residual airflow like you would with a CPU air cooler.

Is your radiator mounted in the top and is it in exhaust or intake configuration?

 

[PhoenixWright]

OK, so I under clocked my CPU to 4.5 GHz which fixed the problem. I did a stress test and I got no errors and the computer didn't freeze but I really want to know why did this fix it? I mean I tried to overclock the CPU Voltage so that each core becomes stable but the computer still froze, does that mean that I had to overclock the CPU and motherboard? I would've been more happy if I didn't under clock my CPU.

Either way, I'm happy with 4.5Ghz and wouldn't mind compromising that little 0.2 for a stable system.

 

[lunyone]

Glad to hear that you got it fixed. It sort of went back to loss of voltage to the CPU that I eluded to earlier, but I'm glad you have a stable system now

 

[Darkbreeze]

Lack of voltage at a given multiplier will cause instability, so you raise the voltage to compensate. This creates more heat. Too much heat creates instability and throttling, mainly from the VRMs but also from the CPU as well if the CPU temps are significantly raised. It's got to be a balance.

You can try going back up by increasing your multiplier by .5 and retesting with the same voltage setting. If you're not also changing voltages during the process then you're missing a step. Always try to use the least amount of voltage necessary to achieve a stable overclock so that electrical stress and temperatures remain as low as possible, while still being enough to sustain stability. Try reading this, it may clear some things up.

Regardless, the 9xxx series chips suck and are no better than the FX-8320, 8350 and 8370. In fact, they are simply the exact same chips, just the best hand picked examples out of the bins, called "binning" or "highest binned" chips, meaning that supposedly even though they are all the same the 8320's will be lower quality chips and the 9590 and 9370's will be the highest binned chips. All they do is put a factory overclock on these chips and pump the voltage way up to keep them stable. This is a terrible gimmick and is why we don't recommend these processors.

My 8320 is currently at 4.5Ghz, and I paid 120.00 for it. Your 9590 is at 4.5Ghz, and you paid 225.00 for it. They are the exact same processor.

Another thing you might want to try is to simply return everything in the bios back to stock settings but turn OFF the turbo core feature so the CPU cannot go past 4.7Ghz and then turn down the voltage in small increments until it becomes unstable. Then bump it back up by .05v or whatever very small amount is required until it again becomes stable. Then one additional small increase and leave it there.

Honestly though, the extra 200mhz is not going to really provide any real benefits worth bothering with unless you're just doing it for the sake of doing it.


http://www.overclock.net/t/91/ultimate-overclocking-guide

 

[PhoenixWright]

Lack of voltage at a given multiplier will cause instability, so you raise the voltage to compensate. This creates more heat. Too much heat creates instability and throttling, mainly from the VRMs but also from the CPU as well if the CPU temps are significantly raised. It's got to be a balance.

You can try going back up by increasing your multiplier by .5 and retesting with the same voltage setting. If you're not also changing voltages during the process then you're missing a step. Always try to use the least amount of voltage necessary to achieve a stable overclock so that electrical stress and temperatures remain as low as possible, while still being enough to sustain stability. Try reading this, it may clear some things up.

Regardless, the 9xxx series chips suck and are no better than the FX-8320, 8350 and 8370. In fact, they are simply the exact same chips, just the best hand picked examples out of the bins, called "binning" or "highest binned" chips, meaning that supposedly even though they are all the same the 8320's will be lower quality chips and the 9590 and 9370's will be the highest binned chips. All they do is put a factory overclock on these chips and pump the voltage way up to keep them stable. This is a terrible gimmick and is why we don't recommend these processors.

My 8320 is currently at 4.5Ghz, and I paid 120.00 for it. Your 9590 is at 4.5Ghz, and you paid 225.00 for it. They are the exact same processor.

Another thing you might want to try is to simply return everything in the bios back to stock settings but turn OFF the turbo core feature so the CPU cannot go past 4.7Ghz and then turn down the voltage in small increments until it becomes unstable. Then bump it back up by .05v or whatever very small amount is required until it again becomes stable. Then one additional small increase and leave it there.

Honestly though, the extra 200mhz is not going to really provide any real benefits worth bothering with unless you're just doing it for the sake of doing it.


http://www.overclock.net/t/91/ultimate-overclocking-guide

Thank you so much for the quick response! I tried all the recommendations you gave me to be able to stabilize my CPU at 4.7 GHz, but I was only able to reach 4.6 GHz with a stabilized CPU.
Anyways, I'm very grateful for you support and If I could give a MILLION best answers I would give it to you.

 

[Darkbreeze]

Hey, no problem. In some cases that's just the way it goes with overclocking. Somtimes you win the silicone lottery, sometimes you don't. I really think that in some cases AMD is pushing out lower binned chips WITH the factory overclock, and calling them FX-9590, and they are simply not able to achieve the clock speeds that are advertised. Personally, I would turn off the turbo features in the bios, cap the clock speed to 4.5Ghz and spec the voltage at about 1.3-1.35v, if it's stable there. Anything above 4.5Ghz is unlikely to really offer much benefit anyhow on these chips and is mostly only for show or to make a point.

It will also run cooler and last far longer there than at the factory overclocked speed and voltage.

 

[PhoenixWright]

Hey, no problem. In some cases that's just the way it goes with overclocking. Somtimes you win the silicone lottery, sometimes you don't. I really think that in some cases AMD is pushing out lower binned chips WITH the factory overclock, and calling them FX-9590, and they are simply not able to achieve the clock speeds that are advertised. Personally, I would turn off the turbo features in the bios, cap the clock speed to 4.5Ghz and spec the voltage at about 1.3-1.35v, if it's stable there. Anything above 4.5Ghz is unlikely to really offer much benefit anyhow on these chips and is mostly only for show or to make a point.

It will also run cooler and last far longer there than at the factory overclocked speed and voltage.

I agree with you, AMD Keeps on pushing their CPU's to the limit. For some people it's useful but it doesn't
help a lot if it's not stable. I think that any person who gets a FX 9590 And can overclock it to 5.0 GHz is a really lucky person.
I heard that it depends on the person's luck(If his card can handle higher clocks).