cancel
Showing results for 
Search instead for 
Did you mean: 

X299-E Random Lockups

skecus
Level 9
Hi Everyone,

I've been experiencing random lockups (with no BSOD) every few days since building my new machine a few weeks ago. When the system locks all the fans and lights remain active but the HD light is no longer blinking and the screen is frozen. This seems to happen during game play or when surfing the web. The only way to get the machine back is to do hard reset with the power button.

Some of the troubleshooting done so far:

-Update BIOS to the latest 0702 version
-Installed latest Chipset Drivers Version 10.1.1.44
-Updated Firmware on 960 EVO and installed the latest NVEe Driver
-Updated latest nVidia drivers
-Checked Windows logs and found no unusual behavior prior to lockups

Other Info - System is liquid cooled running 31c idle and 50-55c on load
I'm running the XPS profile to achieve 3200mhz on the memory, CPU is still stock.

I'm going to try swapping out another power supply, but do not have another board or CPU to test at the moment. Any suggestions or help would be greatly appreciated.


System Specs
Mobo - Asus Rog Strix X299-E
CPU - Intel Core i7-7800X Processor (No OC)
Video - EVGA GeForce GTX 1080 TI FTW3 iCX Hydro Copper GAMING, 11G-P4-6699-KR, 11GB
PS - EVGA SuperNOVA 850 P2, 80+ PLATINUM 850W
OS Drive - Samsung 960 EVO Series - 250GB PCIe NVMe - M.2 (Windows 10x64)
Memory - G.SKILL TridentZ RGB Series 16GB (2 x 8GB) 288-Pin 3600MHz (PC4 28800)
Cooling EK-Supremacy EVO / EK-XRES 140 Revo D5 PWM pump with 2 240M Radiators
7,394 Views
19 REPLIES 19

Chino wrote:
Have you tried running your system at stock defaults?


From the read it looks like he is running stock on CPU, and XMP for the ram.

Try running ram bone stock speed of 2666MHz to see if still having the dead locks (what I call the lockups when there is no BSOD)

I would also run a memtest on the ram, just to be sure it isn't a problem there.

Good luck!

skecus
Level 9
Thank you for the suggestions. I'm going to disable the XMP profile as well as run memtest then report back.

Chino
Level 15
That's a good idea. Clear your CMOS for good measure. Also make sure you're using HCI Memtest because it's the most effective for DDR4 testing.

I've just recently built a system with the X299-E Gaming, initially using an old Toughpower semi-modular 1200W PSU, and eventually switching over to an EVGA 850W P2. The CPU is a 7820X, and the RAM is two sticks of 16GB Kingston HyperX Predator DDR4 3000. Like yourself, I'm running with the XMP profile enabled, but the CPU is currently running stock.

I built the system 4 days ago, and changed power supplies two days ago. Yesterday, I had one hard lock of the system, which required me to power off and on the system. Nothing in the event log so far to indicate what happened, and nothing of note really occured when the system froze. It wasn't under load at all, and the temperatures weren't an issue.

I'm going to leave the system as is for a while, and see if it comes back again. I'd be interested to know if switching off XMP helped for you.

I had similar issues with my R4E build. Random glitches/lock-ups that were hard to trace to a specific source

In my case, it was the RAM. Specifically, one of the secondary timings. Though I don't recall which one, one of them had to relaxed quite a bit (maybe tRFC?)

From what I have heard, XMP only sets the primary timings, not secondary or tertiary

You could check out this post from raja about RAM timings on the Crosshair V:
https://rog.asus.com/forum/showthread.php?2585-ASUS-Crosshair-V-Formula-BIOS-Guide-Overclocking&p=15...

Well, I've managed to find something that would cause the system to run unstable enough for an application crash or complete system freeze. Funnily enough, it's Batman Arkham Knight.

So, setting every graphic setting as high it could go, enabling all of the nVIDIA Gameworks settings and running at 3440x1440 at 100Hz with GSync enabled, I could either force it to cause an application crash or total system lockup when running the XMP profile of my RAM. Interestingly enough, setting it back to stock would enable the game to run glitch free. I'm about to set things back to XMP to see it isn't just coincidence.

I haven't run Arkham Knight before I swapped out the old Thermaltake 1200W to the new shiny EVGA 850W P2, but I didn't experience any issues running 3DMark and a couple of other games when using XMP with the Toughpower.

I did have an issue trying to run the USB3 hub my monitor provides when using the Toughpower (it'd always report that a USB device failed, regardless of what port I'd use), but I no longer have that issue when I use the EVGA. As for USB devices, I have a Logitech gaming keyboard and mouse, 720P webcam, Steam controller dongle, Xbox One controller dongle, Logitech wireless headset dongle, sometimes a Saitek X52 HOTAS (only plugged in when I use it), the monitor's USB3 hub, and an internal USB hub as I have front port USB 2.0 ports and an NZXT cooler, but only one USB2 header on the motherboard.

So... in summary:

Random lockups seem to be related to XMP, and running stock seems to fix stability. I'm about to reboot and verify that.

System seemed more stable when running XMP and a different power supply. Again, I'll need to verify that. The power supply used could be a red herring.

Other than the wattage outputted, the main difference between the two power supplies is the EVGA has a single beefy +12V rail (70.8A, 849.6W) and the older 1200W Toughpower has multiple +12V rails (20A on the first +12V, 36.0A on the second +12V, 30A on the +3.3V, 600W combined for the lot, and the same for the second set).

Edit: Okay, can confirm that setting the XMP profile causes Batman Arkham Knight to crash. What little of it that runs also seems to indicate that there's some performance loss as well.

I'll run a few other things to see if anything else causes issues as well. If I get a chance to, I might also see how it goes when using the old Toughpower later this week or next weekend.

Chino
Level 15
The best way to see if your problem is caused by memory instability is by testing the memory with HCI Memtest or Google Stressapptest for Linux.

skecus
Level 9
Sorry it's been a few days, I had to leave town for business. Last night I ran HCI Memtest on the available memory. Each time I ran the tests, the system would lock up anywhere from 2-20 minutes. Since I recently purchased the Gskill memory on Amazon, I was able to get a full refund. I decided to go with Corsair Vengeance LPX DDR4 4000. As soon as I get these installed I will run some stress tests and report back.