Motherboard MSI SLI PLUS X399
Processor AMD Ryzen Threadripper 1950X
Cooler Arctic Freezer 33 TR CPU
Graphics Card MSI GeForce GTX1050 Ti 4GT OC 4GB
Power supply Seasonic FOCUS PLUS Modular (80 + Gold)
SSD Corsair Force MP300 M.2 SSD PCIe NVMe 480GB
RAM 2 set of Kingston HyperX Fury black 32GB DDR4 K2 HX426C16FBK2
My computer hangs under load and the monitor turns off. The BIOS is updated, the Video card is inserted into another slot, I tried it with another video card, the drivers are all updated. Conducted stress test AIDA64, the computer kept the temperature in the normal range, after an hour of the test, the screen went out and the system hung. All equipment is new from the store. For several weeks I have been trying to find the cause and nothing comes out.
Sorry for bad english writing via google translator.
I will be very grateful if you help solve this problem.
idiscodancer, I think I am reading your English just fine. What is the power of your power supply? I have the same processor with an 850 Watt supply. Any less may be a problem. Tell us your UEFI and OS versions. Please post a screenshot of Ryzen Master (RM) with your system under load. Have you made any modifications in your UEFI (XMP)? Please try a Clear CMOS and run again. Thanks and enjoy, John.
Hi, John Thanks for the answer! Power supply 850 W, sorry forgot to write. The installed OS Windows 10 (1809) version of UEFI 7B09vA6. The system is not overclocked, all BIOS settings are by default.
Thanks, idiscodancer, this all looks good. Please post the RM screenshot I requested above. What power plan are you running. Please make sure your chip set drivers are from AMD. Here are your specifications:
I am confused on your memory - four sticks of 16 GB for a total of 64 GB? I need to see your RM screenshot. Thanks and enjoy, John.
idiscodancer, I am very concerned about your CPU temperature. I hate to suggest for you to change your cooler, but air is really not enough for these big processors and water is almost mandatory. I use a Enermax Liqtech TR4 240 AIO Liquid CPU Cooler, which cost me about USD 140. It may be really expensive for you and you must consider fitting into your case. There are many available which are less expensive, maybe more so for you. Please be sure that the water block covers both your chips. I still do not know how much memory you have - PLEASE tell me. Also please run the free, read only, version of Thaiphoon Burner and post a screenshot. I will not post a link because it causes my reply to be moderated delaying it a few minutes to a day or more. Please search "Thaiphoon Burner" and DL the free version. I am concerned about the memory speed/size. You may need to slow your memory or up your SoC voltage. Until you decide on your CPU cooler, please do not increase the SoC voltage. Enjoy, John.
Hi john Thanks again for your advice. I took my computer to the service center, the experts found that the BIOS was updated with an error plus, I had an old hard disk for data storage, which had a lot of errors. After replacing the hard drive and updating the BIOS, the problem began to arise much later, not after 1-2 hours under load, but after 15 hours. But the problem still remained. Ustnovil Thaiphoon Burner, I hope this is what you asked for. I have 64 GB of RAM, 4 dies of 16 Gb each. The service also noticed that one of the memory kits turned out to be 2400 GHz, instead of 2666, apparently the Kingston plant was not properly staffed.
Also attach the Windows event log, confused error 56 Application Popup. Kernel-power 41 error occurs due to restarting the computer from the button after a hang.
Thanks, idiscodancer. I usually do not pay much attention to the event log and in Cirillic (?) I can't make much sense of it. The real problem found seems to be the mismatched memory. The Thaiphoon Burner data looks the same for all sticks. Was this after it was corrected? What do you think about an AIO water cooler? Your RM screenshot indicates you are running at 2400 MHz. This must the the SPD speed and I suggest you leave it there, till this is resolved. Going to 2666 MHz will only make it worse. What memory slots are you using? Thanks and enjoy, John.
EDIT: What time zone are you in? I'm UTC - 5 hours right now.
Do you think that the problem could be in the processor overheating ??? The temperature of the loaded processor is not higher than 66 degrees, and in the quiescent state it is not higher than 43-44 degrees. Could the problem be a motherboard malfunction?
Here is the AIDA64 test data.
After the test, I increased the speed of the processor fan and the temperature did not rise above 64 during the AIDA64 stress test
idiscodancer, the problem could definitely be the MB. I had a hang with similar symptoms on my 1800X but not under load, just idle. That turned out to be a bad memory slot and required an MB RMA. You could test by plugging your memory into the other four slots. I do not think you ever told me what slots you were using, but it is usually A2, B2, C2 and D2. A1, B1, C1 and D1 should work also. Which every four you are using, please try the other four. It also could be a memory stick. You could try running with one stick at a time and see if you see a failure. I am assuming the mismatched sticks in the quad kit has been corrected. You should always run your CPU fan at full speed unless it is too noisy. One more advantage of water. Enjoy, John.
John, is it possible to put the RAM in the slot B2 ?? In the instructions for the motherboard it is written that the first strap of the RAM should be in slot B2. I also noticed that periodically all the fans of my computer turn on by themselves, even when idle. Does this indicate a malfunction of the motherboard?
Hi john I installed an additional fan on the CPU cooling radiator, now the temperature is under load of 58-59 degrees Celsius. I tested every level of RAM in slot B2, the computer worked stably and did not hang. But when you try to install two memory bars in slots B2 and D2, the computer freezes again.
Great, idiscodancer. Sounds like you are own the trail! I assume you have tried all sticks in D2, one at a time, you get hang? If so sounds like you need to talk to your MB vendor about RMAing the board. You can put memory sticks in any slot, at least on my MSI board. The UEFI complains the memory location is not optimal, but it runs fine. Great work! Looks like you have solved your temperature problem and can use your air cooler. Please keep me informed - looks like you are getting close. Enjoy, John.
Hello John! THE PROBLEM IS SOLVED !!! The whole thing turned out to be in the unfinished BIOS from MSI. Changed in the parameter "Global C-state control" from auto to disable and the computer works and does not freeze !!! Only a month of my life I spent on experimentation to solve my problem and not one of the technical support did not help me! I just heard that replace the unit, try your hardware somewhere with a different build !!! I have already put my RAM on sale. But in one of the forums I found the answer, where many had a similar problem. But it manifested itself when overclocking the processor. But I have everything at custom frequencies !!! I am wildly outraged by this! I will never buy MSI products again !!! Did you know about such a problem?
I concur your CPU temperature is quite high at idle (3.45 ghz) @ 64c.
You need to recheck your installation of the Arctic Freezer 33 to make sure it is making a tight contact with the Ryzen surface. Very carefully, check to be sure It doesn't wobble or be able to move around on the CPU. Make sure that the base of the Arctic is completely making contact with the CPU surface.
If you put too much Thermal paste it could prevent proper cooling by the Cooler. Best way is to spread the paste thinly with a credit card or flat wooden stick over the entire contact area of the CPU Cooler copper surface. When you reapply the thermal Paste make sure to carefully clean all previous Thermal Paste with isopropyl Alcohol and a cloth or cotton ball from both the CPU and Cooler surfaces.
As a troubleshooting test, Remove one side of the Computer case and then check the Ryzen CPU temperature at idle. If it goes below 60c at idle it could indicate poor circulation in the Computer case. If it continues to go above 50c at idle it could indicate a improper installation of the CPU cooler or you need an extra fan or stronger CPU Cooler.
Try using your motherboard's BIOS setting for Fan Control to maximize the Fan Speed to keep the temperature lower at idle.
It may be advisable to add a second Fan to the Arctic Freeze 33 cooler in a PUSH-PULL arrangement. This will help when the CPU is under heavy loads.
I have a FX8350 which is a 125 watt CPU and I recently installed a CoolerMaster Hybrid 212 EVO CPU Cooler with two fans (Push-Pull). At idle my CPU is in the Mid 30c and at very heavy loads it doesn't go above 59c.
Here is a previous AMD Forum thread about having high idle temperatures on a 1950x CPU: Ryzen Threadripper 1950x dropping clock speed
It mentions ways to lower the CPU temperature through Ryzen Master Power plan and rechecking the CPU Cooler installation.
No, you did not understand correctly, on the screenshot, the temperature of the processor is in a loaded state. At rest, its temperature is not higher than 45-46. With loads usually 64-66 degrees. The second fan has already ordered the Arctic Freeze 33
Adding a second fan will probably help a little bit in keeping your temperatures down. Also use BIOS to configure the FAN to run more aggressively during lower temperatures to keep the CPU cooler.
Okay, I let Misterj continue troubleshooting your problem.
Good luck
thanks for the answer