Prime95 and Aida64 crash reboot

[visualfx]

Hello,

I have problems after buying some hardware. Prime95 rebooting without any error messages after 3-5 hours, tested with torturetest with small fft and Aida64 rebooting after 6-8 hours tested with stresstest FPU checked, CPU and Cache tests seems stable. Memory tests on Windows 7 desktop with memtest and Aida64 memory checked rebooting without error messages but memorytest on bios with memtest86+ and windows 7 built in memorytester(advanced options) seems stable. I didnt overclocked and my cpu is also not made for overclocking anyway I didnt change the standard Voltages either. My CPU core voltage shown on Aida64 is 1.056 V but with CPUID HWMonitor I read max 1.121 V. I have searches the net and found out that is may occur due to heat problems or due to lack of Powersupply.

First to heat I read max 76 C with Aida64 and max 81 C with Prime95 and Powersupply is a brand new Powersupply with 600 Watt. The Powersupply should be enough and my processor cooled with standard (came with the cpu) airfan should easily take 90 C.
The temperatures I read with HWMonitor I guess this is tjunction and not tcase
İs there any programm out there to read tcase temps ?
tcase max is 67 C for this cpu but I dont know the max tjunction for my cpu
The Hardware I bought:

CPU i5 3570 standard airfan cooling not overclocked

Motherboard Gigabyte GA B75M-HD3 none of the voltages changed

Ram 2x DIMM1and DIMM2: Kingston HyperX KHX1600C9D3/4GX 4 GB DDR3-1600 DDR3 SDRAM (11-11-11-28 @ 800 MHz) (10-10-10-27 @ 761 MHz) (9-9-9-24 @ 685 MHz) (8-8-8-22 @ 609 MHz) (7-7-7-19 @ 533 MHz) (6-6-6-16 @ 457 MHz)

Powersupply Corsair CX600M 600 Watt.

Im running Windows7

and I have Sapphire HD5770 graphic card.

so what you think is the problem my system not being stable ?

thanks

 

[Ed_1]

what timings are you running , 1600 or 1333 ?

http://www.kingston.com/datasheets/KHX1600C9D3_4G.pdf

looks like you need Vram =1.65v for 1600 .

Have you tried one stick at time to see if one is bad (this needs to be single channel mod slot) .

 

[visualfx]

Supported Module Timing at 800.0 MHz: 11-11-11-28
Module Nominal Voltage (VDD): 1.5 V
if you run with a latency 9-9-9 then you need 1.65 V but not with 11-11-11
As I mentioned the ram tests with memtest86+ and windows 7 built in memory checker showed no errors. And I dont have single channel mod slot
What You think about my Temperatures

 

[Ed_1]

temps are fine for stock cooler .
FWIW :my 3570k at stock clocks with prime95 does about 62-65c with CM 212evo HS , I never installed retail HS to compare .

try bump Vram to 1.55v

have you doubled checked all power plugs (24pin and ATX 8pin ) , I always make sure each pin is all way in .

Any min voltages much below spec like 12v rail while load is on ?
Also if you run prime95 with large FTT doe it take longer or shorter time to crash, or not crash .

Large generally pushes memory more and small more CPU and L1/2/3

 

[kn00tcn]

are you sure it's rebooting without error messages? cuz windows is set by default to NOT show you the BSOD

 

[visualfx]

powerplugs checked.
min voltages not checked I will do it tomorrow
large ftt test I will do it tomorrow
Where can I set messages to show in windows when It reboot?
thanks

 

[Ed_1]

powerplugs checked.
min voltages not checked I will do it tomorrow
large ftt test I will do it tomorrow
Where can I set messages to show in windows when It reboot?
thanks

control panel>system>advanced system settings>advanced tab>startup and recovery settings , untick automatically restart .

Blue screen viewer may help if you got them to see whats cause .
http://download.cnet.com/BlueScreenView/3000-2094_4-10965136.html

 

[Jetster]

Fill out your System specs in the user CP. All your specs. You don't need to run Prime for 3 hours. Any system will fail if you run it long enough. Temps above 80c are not normal. Run everything stock try it for a couple of days if you get no restarts then run your memory at xmp memory profile. After doing what Ed_1 said shutting off automatically restart

 

[Ed_1]

Fill out your System specs in the user CP. All your specs. You don't need to run Prime for 3 hours. Any system will fail if you run it long enough. Temps above 80c are not normal. Run everything stock try it for a couple of days if you get no restarts then run your memory at xmp memory profile. After doing what Ed_1 said shutting off automatically restart

http://www.techpowerup.com/reviews/Intel/Core_i5_3570K_and_i7_3770K_Comparison/10.html

 

[Jetster]

Everyone knows that Ive runs hot. He needs to push more air past it or change the heatsink. It can cause errors

 

[visualfx]

Hello,
I run today Prime with large FTT it crashed after 15 minutes.
voltages for +12 V 8.880 and for -12V -4.224
I run memtest (on windows) it showed no errors till 115% coverage then I went to the mall when I came back it had crashed already and windows rebootet before I was home, but I mentioned earlier in my post memorytest with windows built in memorytester and memtest86+ (bootet with usbstick) showed no errors.
I really confused because some people say it is temperature and some say it cant be temperature. I have posted same thread on seven different websites and there is the same some say temp others not, some say PSU.
Does anybody know exactly what is the max temp for my CPU. I mean not tcase because I dont know how to measure tcase I have HWMonitor and I guess it measures tjunction. İt has temp values for the four cores. I suppose this is tjunction and it has a measure called package dont know what this is for.
thanks

 

[vega22]

package is tjmax.

8v on a 12v line?

i think you found the issue if it aint a misread


tjm is 105c on ivy afaik btw dude

 

[visualfx]

Hello,
I have posted an image so you can look.
but I didnt change anything on my motherboard settings. I even dont know how to do that.
İs it possible that the standard settings are incorrekt so give me a hint how can I correct that and primaly with which programm
and what should the setting be.
thanks

 

[vega22]

that has to be a misread.

i mean if your 12v was only 8v it would not work.

 

[Ed_1]

Hello,
I have posted an image so you can look.
but I didnt change anything on my motherboard settings. I even dont know how to do that.
İs it possible that the standard settings are incorrekt so give me a hint how can I correct that and primaly with which programm
and what should the setting be.
thanks

well 8.8v is way bad, so bad I am not sure its reading right voltage . At idle is it also way below 12v ?

You wouldn't have voltmeter to test out on the 4 pin plug ,12v wires ?

12v line shouldn't go below like 11.5 v under load .

See what HWinfo64 gives, if it matches HWmonitor , also see what bios reports on voltage as that one should be correct .

 

[Jetster]

First off. The stock CPU cooler is not for stress testing. Gaming will not push your CPU that hot. So forget stress testing with the stock Cooler. Second your have contradicted yourself about Memtest86. Run two passes. If it with no errors then your good. Then just use the PC. If you get blue screens then something is wrong. Shut off auto restart and it will stay on the BSOD when it happens. Quite messing with voltages set everything STD

And stop using HWmonitor. Use realtemp

 

[visualfx]

Hello,
+12V with HWMonitor was a misread I see now 12.024V with HWİnfo I run again a Prime with large FTT Max Coretemp with HWİnfo was 74 C.
İt crashed after 10 minutes. I will run memtest86+ now, so I wont be able to post as the test will take several hours see you tomorrow.
By the way I had problems with games therefore I strated stresstesting. Actually three games averagely crashes after 2-10 minutes. and 3dmark also crashing.
All this after I built this new system the only hardware I didnt change was my graphic card which I will change january to a ASUS HD 7950.
thanks

 

[Ed_1]

that voltage and temps is fine for prime95 large FTT .

Try using lower 1333 for ram , leave lose timings .

 

[visualfx]

Hello Im here again. I did several tests here are the results.

1 Test memtest86+
3. pass reboot without error messages in the programm

2 Test memtest86+
5. pass reboot without error messages in the programm

3 Test memtest86+
4. pass reboot without error messages in the programm
then I thought the ramms fine just having issues running stable for long time

4 Test Prime95 large fft
1 minutes reboot

5 Test Prime95 larfe fft
1 minutes reboot

6 Test Prime95 small ftt
2.5 hours reboot

7 Test Aida64 cache stress test
4 hours stable no errors then I stopped the test

this is very strange if it is not the cache and ramms are running in one test 5 passes
why the hell Prime95 large fft reboot after 1 minutes.
I guess some kind of devil is inside my system, I need an exorcist

thanks

 

[Ed_1]

Hello Im here again. I did several tests here are the results.

1 Test memtest86+
3. pass reboot without error messages in the programm

2 Test memtest86+
5. pass reboot without error messages in the programm

3 Test memtest86+
4. pass reboot without error messages in the programm
then I thought the ramms fine just having issues running stable for long time

4 Test Prime95 large fft
1 minutes reboot

5 Test Prime95 larfe fft
1 minutes reboot

6 Test Prime95 small ftt
2.5 hours reboot

7 Test Aida64 cache stress test
4 hours stable no errors then I stopped the test

this is very strange if it is not the cache and ramms are running in one test 5 passes
why the hell Prime95 large fft reboot after 1 minutes.
I guess some kind of devil is inside my system, I need an exorcist

thanks

did you change windows settings as my above post .
control panel>system>advanced system settings>advanced tab>startup and recovery settings , untick automatically restart .

Blue screen viewer may help if you got them to see whats cause .
http://download.cnet.com/BlueScreenView/3000-2094_4-10965136.html

If you did and it still reboots w/o error , BSOD then I would try another PSU if you can borrow one for test .
another thing you could try, go back to basic's ,remove everything so you have just MB/CPU/Ram/PSU and HD (you could even run w/o ATI card , using HD4000 Intel ) , out of case and see if same thing happens .

 

[visualfx]

untick automatically restart I did now
BlueScreenview I already installed 4 days ago it shows no dump file after several crashes
I will tomorrow arrange another PSU but only 500 Watt I hope it will be enough to test.
see you tomorrow...

 

[Ed_1]

untick automatically restart I did now
BlueScreenview I already installed 4 days ago it shows no dump file after several crashes
I will tomorrow arrange another PSU but only 500 Watt I hope it will be enough to test.
see you tomorrow...

If you only unticked now, I would try another prime95 test as it only takes a min to error/reboot .
If you get info from it it might save time .

 

[visualfx]

I did, Prime95 reboot after 8 minutes, but no info
and BlueScreenview also shows nothing

 

[Ed_1]

I did, Prime95 reboot after 8 minutes, but no info
and BlueScreenview also shows nothing

Ok, thats pretty fast .
I would try another PS as that's fairly easy to swap out an test , it sounds like power issue but you never know . Just make sure all power connections are firmly in all the way . With new MB many times there tight fit so might not be home all the way . You got 24 pin and 8 pin EATX plugs

 

[visualfx]

Hello again,
it was PSU. I installed an Enermax 500 Watt all test passed. Prime95 large ftt 13 hours stable. 3D Mark works. then I plugged the old PSU back.
to my surprise It was also stable. This means the power connectors wasnt plugged correctly in first time but I swear I did checked twice so I dont know what it was but now everthing works as it should. İn 2 Days Ill get a new case and built the rig again we will see what happens then.
Thanks to you guys who helped during this difficult situation.

I have another question next month I will upgrade my gpu.
I was thinking between R9 270X GTX 760 and HD 7950. My choice was a ASUS HD 7950 DC II Top.
because of 3 GB VRAM and the higher 384 Bit. My İnvestment has a time goal of 3 years. I dont know if coming games requıre more ram or a better chip. What do you think of this choice.
Thanks