General Discussions

[wexy123]

Hi. I have a question that someone might be able to answer me.

I got a new video card and after using it for a couple of hours problems happened while I was playing games. So basicly when I start exact games (for example Rocket League, Hell Let Loose, SWBF2) after a couple of minutes of play my pc keeps shutting down (like when the power is gone) and reboot immediately. This never happened with my old card so I'm guessing everything is fine with my rig. Obviously I have the latest drivers, I tried reinstalling windows, cooling is not an issue because my GPUs max temp is around ~65C everytime and the CPU and the motherboard is max ~50-60C also my PSU is using around 120-130W max, cables are not touching any compontent in the pc's case, I tried stress tests both on the CPU and GPU and everything was perfect I got no shut downs, I tried reconfigure voltages in Radeon Software but the problem still occurs. I somehow figured out if I play on windowed mode or medium settings it solves the problem for a while. It's really annoying so I accept any possible solution. Thanks!

My Specs are:

Motherboard: Asus P8Z68 Deluxe
Ram: Corsair Vengeance 4gb DDR3 1600MHz
GPU: MSI Radeon Rx 480 8Gb GDDR5 265bit (used to have some Amd 3000 series 2Gb I'm not sure which lol)
CPU: Intel Core i7 2600k 3.4GHz
PSU: Chieftec iArena 600W
HDD: Western Digital 1TB
Os: Win 10 64bit
CPU Cooler: Cooler Master Hyper H412R

[Anonymous]

With peaks over 150 watts of power consumption of your graphics card, your power supply might no longer be sufficient at full load. The 6-pin PCI Express connector only provides a maximum of 150 watts. Carry out the installation of the Radeon driver 21.12.1 as follows: Download the graphics driver and chipset driver of the motherboard from the AMD support page, install the chipset driver and update Windows Update, then deactivate the updates, set the device installation setting to no in Hardware (prevents the installation of a Microsoft driver for the Graphics card that does not get along with the Radeon), download the graphics tool DDU, deactivate the network card, delete everything in safe mode with DDU and restart, install the Radeon driver, manufacturer display driver possibly, restart again and now activate the network card. If your problems are not solved after this procedure, I would consider a new power supply with approx. 750 watts, which is also suitable for a new PC, preferably Gold +.

[Anonymous]

The problem is also that the Radeon driver is constantly optimizing itself, i.e. not acting statically. If then the power consumption is exceeded, crash and restart because of WattMan. It's similar with my RX5700XT. That's why I don't use the adrenaline software that does what it wants, not what I set, but the MSI Afterburner. It is much easier to handle, you can work well with 5 graphic profiles, in addition to temperature limitation, core clock and core voltage.

[elstaci · Accepted solution]

download and run OCCT PSU Test and see if your PC shuts down.

While running the PSU Test keep an eye on your PSU outputs of 12/5.0/3.3 Vdc. They all should be within 5% -/+ tolerance.

For instance -5% of 12 Vdc is 11.4 Vdc. So your 12 Vdc should not go below 11.4 while being tested.

Also keep an eye on both the GPU and CPU temperatures and Fan speeds.

What was the Make & Model of your previous GPU card that you said works fine in your PC.

Also post an image of GPU-Z  to make sure your AMD Driver is correctly installed and to verify your new GPU card is authentic and not a fake.

NOTE: According to MSI Support you need  a minimum PSU of 500 Watts so your PSU is more than enough to run your GPU card in your PC. But it could be defective and not giving out proper PSU voltages under heavy loads.

From what I have read about the reviews on your Chieftec PSU is an inexpensive (cheap) PSU so it is possible under load it is not providing the proper voltages to run your GPU card or PC as a whole while under heavy loads like gaming. just guessing though.

[wexy123]

I tried out OCCT PSU test and after I pressed the start button my pc immidiately shutted down. 

Here is the screenshot: https://ibb.co/8PkX6MY

[elstaci]

Okay, Thanks for the image.

By GPU-Z, it looks like your AMD Driver is installed correctly and all API's correctly checkmarked at the bottom.

Try running just OCCT GPU Test MEMTEST and see if it shuts down. If that passes then run the GPU Test and see if the same thing occurs.

If the tests doesn't shut down your computer take a look at the PSU outputs and temperatures and Fan speeds while testing.

When OCCT shuts down your computer during the  testing so quickly, in my opinion, indicates hardware issue.  Either RAM, GPU, or CPU. Like Overheating, improper Overclocking, PSU power issue, etc.

If you install the GPU card in another computer and run the same test and the computer also shuts down in another PC then you have a problem with your GPU Card.

[wexy123]

*Update* 

I bought a new PSU and now with the OCCT stress test everything seems fine. I got no crashes so I guess the PSU was faulty. Thank you for everyone for the solutions.

[witty245]

when you changed your gpu did you swap it out with an identical model as if you upgraded it you may find new gpu needs more power, also you may need to consider how old is your psu as if its over 5 years old it may have degraded and not able to produce optimal power like when new. also only consider good makes with psu sesonic etc another thing that can cause system instability is your ram settings for instance my ram is rated at 4000mhz but if i run at 4000mhz i will get random system restart but at 3800mhz its totally stable

another thing you could check is the gpu overclocked if yes set it to default settings

[kingfish]

Did you do a clean install, using DDU in safe mode, to remove all vestiges of the previous drivers? If your previous card was Nvidia, then run DDU again except choose Nvidia drivers. Reboot and install the correct AMD driver.

Do you have your computer power plan set to high performance..not the default 'balanced'?

[KRAVER]

I have same/similar problem.

Threadripper 1920x CPU
x399 strix MOBO
strix vega64 GPU
64Gb gskill 3600 c16 (2x 16gb neo, 4x 8gb)
Corsair HX1200 PSU
EKWB CPU cooled with 2x 280se rads, Pump/res D5

CPU vcore 1.225 undervolted
CPU runs 4.0ghz all core
ram runs stable @ 2800 16-18-18-18-36 1.35v T1

this PC was never overheated.
CPU while running P95 is 65c max
GPU while running R20 or R23 is 70-75c max

Problem happens while gaming only. its like someone hit the reset power button. PC reboots and back to windows. 
I reinstalled windows over a dozen times now and still does it. been trying to figure this out now for weeks.
did memtest passed 2x, tried 1 ram stick at a time with all 6 sticks no issues. tried different GPUs same problem.
checked my PSU with a PSU checker no issues. all voltages are perfect.
originally thought my google account was corrupted and was messing with my PC but it was not. same problem still happens.

I even setup a basic windows 10 x64 Pro with no extras.
didn't sign into google account.
only drivers and basic programs needed to run PC.
(afterburner, AI suite 3, CCC 21.12.1 )

worked for a few hours and back to resetting while gaming.

Im pretty good with PCs been working and fixing them for 25yrs now so I usually dont get stumped but I am now.
any help would be greatly appreciated.
this exact setup ran for over a year with no problems till Dec 2021

[elstaci]

@KRAVER 

Just for your information only your ThreadRipper Maximum Operating Temperature is 68c. Which means once the temperature starts reaching or reaches 68c the processor starts to throttle automatically to keep the temperature at 68c or below.

but if you tested it by stress testing the CPU and the highest temperature it got was 65c that is still within its Maximum Operating temperature.

Try running OCCT PSU Test which runs both GPU and CPU tests at the same time putting maximum wattage to the PSU. Keep an eye on Temperatures, Fan speeds, and PSU Outputs which should be within 5% tolerances.

Recently a User with a Corsair 1200 watt PSU found out that the PSU was not 100% compatible with his motherboard which Corsair mentioned as a flaw with those models.

The User replaced his Corsair 1200 which was good but shutting down his computer randomly with a Corsair 1000 watt PSU and that fixed his problems. The User mentioned that Corsair Support said the 1200 watt unit can't handle certain voltages on certain motherboards causing the PC to shut down.

The User ended up using the 1200 watt Corsair PSU on another computer without issues.

I would open a Ticket to Corsair Support and have them verify that the 1200 Watt PSU isn't compatible in all motherboards.

NOTE: I tried to locate the AMD Thread where the User above found out the problem was the PSU but couldn't find it.

EDIT: Found a Tom's  Hardware article from 2021 where Corsair has issued a Recall for some HX 1200 watt PSUs: https://www.tomshardware.com/news/corsair-issues-hx1200-hx1200i-psu-recall

Corsair says that some HX1200 and HX1200i units may cause compatibility issues with certain motherboards, but the company didn't go into finer details. Essentially, the power supply will prevent the system from booting. The issue may arise immediately or after installation. Corsair has emphasized that the problem will not damage your hardware, but you should still call in for a replacement unit anyway.

The impacted HX1200 and HX1200i power supplies come with lot codes from 2030XXXX to 2041XXXX, which were typically commercialized after July 20, 2020. Corsair has provided a short and simple guide to decipher the lot code for your unit.