PC crashing in games

[Atriox]

PC Specs: RTX 4090 STRIX OC, 13900K, x4 Kingston 16GB Ram 6000 MT/s, Corsair HX1500i

I tested my GPU in Heaven Benchmark and it didn’t have any crashes.

I tried to see if I could run unreal engine 5 shader caches and my CPU isn’t crashing.

I am wondering if it is my ram responsible - I tried multiple times to try and test the ram but it never lets me. Whenever I try and test it, I have no signal on my monitor, and I also never have signal when trying to enter safe mode.

Does anyone have any idea how to figure out if it’s the RAM if I can’t get a signal on my monitor when trying to test it? Using different cables doesn’t help and there’s no issues with the monitor.

I just 10 minutes ago had a new issue where my entire PC froze…

[_I_]

post a cpuz validation link
http://www.cpuid.com/softwares/cpu-z.html
cpuz -> validate button -> submit button
it will open a browser, copy the url (address) and paste it here

compare the spd tab (jdec/xmp profiles supported by each dimm) to memory tab (what its running at)

if you are mixing ram, or different kits, you cannot use xmp, and will need to find compatible speed/timings/voltage for all dimms

[Atriox]

_I_ a écrit :

post a cpuz validation link
http://www.cpuid.com/softwares/cpu-z.html
cpuz -> validate button -> submit button
it will open a browser, copy the url (address) and paste it here

compare the spd tab (jdec/xmp profiles supported by each dimm) to memory tab (what its running at)

if you are mixing ram, or different kits, you cannot use xmp, and will need to find compatible speed/timings/voltage for all dimms

The ram is all the same and it supports 6000MT/s when used in XMP - I set XMP in the BIOS and it’s running at 6000

[_I_]

ok, but still, post cpuz validation url
and make sure its running at its rated speed/timings for the 6000 profile

also check the fsb, some intel cpus dont like being over 100, 99.9 is fine

in bios, after you set xmp, you can try bringing ram speed one bump, 5800
see if thats stable to rule it out

what gpu power cables are you using?
any adapters?
HX1500i does not appear to have the 12vhpwr connector

yea update bios if you have not, cpuz validation would also show that
also grab the drivers from the mobo site, not windows update

[GOD RAYS ON ULTRA™]

https://www.theverge.com/2024/7/26/24206529/intel-13th-14th-gen-crashing-instability-cpu-voltage-q-a

Are you aware of the stability issues with the 13900k? The situation is rather nasty.

https://www.theverge.com/2024/7/26/24206529/intel-13th-14th-gen-crashing-instability-cpu-voltage-q-a

It's all over the internet and youtube. Those are just the first examples. You're supposed to update the bios, use intel defaults, and pray there's no permanent damage.

[Atriox]

_I_ a écrit :

ok, but still, post cpuz validation url
and make sure its running at its rated speed/timings for the 6000 profile

also check the fsb, some intel cpus dont like being over 100, 99.9 is fine

in bios, after you set xmp, you can try bringing ram speed one bump, 5800
see if thats stable to rule it out

what gpu power cables are you using?
any adapters?
HX1500i does not appear to have the 12vhpwr connector

yea update bios if you have not, cpuz validation would also show that
also grab the drivers from the mobo site, not windows update

https://valid.x86.fr/msscnu

I am using a GPU adaptor that came with the RTX 4090 STRIX OC

[Atriox]

GOD RAYS ON ULTRA™ a écrit :

https://www.theverge.com/2024/7/26/24206529/intel-13th-14th-gen-crashing-instability-cpu-voltage-q-a

Are you aware of the stability issues with the 13900k? The situation is rather nasty.

https://www.theverge.com/2024/7/26/24206529/intel-13th-14th-gen-crashing-instability-cpu-voltage-q-a

It's all over the internet and youtube. Those are just the first examples. You're supposed to update the bios, use intel defaults, and pray there's no permanent damage.

I'm fully aware of those issues and the CPU is new - I RMA'd it early September.

[BurakZG]

There is no way you could reliably test RAM from any OS. The OS (does not matter, Windows or Linux) is limiting access to the memory to apps. That is normal.
You have to make bootable pendrive with memtest86. The same memory test is included in any bootable linux installer. So you could also download Ubuntu or Manjaro, make bootable USB stick and you can test memory.
It's actually advised to test RAM after any change to the RAM configuration in UEFI. Especially if you observe random reboots.

[_I_]

4 dimms is more load on the cpus memory controller
try bumping it down a few steps, if that make it stable, then bring it up one at a time

6000 cl40 is not very fast, that puts it near generic/oem ram speeds

you can grab the usb bootable memtest86+ to test ram
or windows memory test tool from recovery boot options

[SimicEngineer]

Losing signal when booting safe mode or memtest86 might be caused by the output switching to the iGPU. Have you tried connecting the video cable to the motherboard when this happens?

[Atriox]

SimicEngineer a écrit :

Losing signal when booting safe mode or memtest86 might be caused by the output switching to the iGPU. Have you tried connecting the video cable to the motherboard when this happens?

Yep, still no signal

[Atriox]

_I_ a écrit :

4 dimms is more load on the cpus memory controller
try bumping it down a few steps, if that make it stable, then bring it up one at a time

6000 cl40 is not very fast, that puts it near generic/oem ram speeds

you can grab the usb bootable memtest86+ to test ram
or windows memory test tool from recovery boot options

I don’t think this would help - I tried running it with XMP off and still had issues

[Illusion of Progress]

My first troubleshooting angle would be to try with XMP off, because you're attempting to run four DIMMs at a pretty high speed (you'd have been better off with 2x 32 GB if you wanted that capacity at those speeds). Four DIMMs, especially with DDR5, incurs a significant penalty in the maximum speed you can stabilize.

You say that still has issues with XMP off, so my next attempt would be to try removing half the RAM (and testing in both pairs).

I'm not sure why you can't test the RAM? MemTest86 or Windows Memory Diagnostic doesn't work? If it's getting unstable, and always, and soon as you go to test it, that's odd, and maybe suggestive as to where the issue lays.

Unfortunately, the CPU itself may be in question simply due to what it is. You said you tested it with shader compilation moments, and that's a good guess since that is a condition which tends to trip them up, but it might not always. Some of the problematic CPU saw stability return when the RAM speed was dropped and the core speed boosting limit was dropped. I'd try this again with a p-core limit set really low (54, sometimes lower) and XMP off (or at least RAM at 4,400 MHz or lower) to see if it improves stability. I see you already did an RMA, but others have had subsequent failures, sometimes two or three of them. If you updated the BIOS though, I would think the chances of that should be low.

[_I_]

yea, try taking out 2 dimms and see if that helps

[BurakZG]

why? he didn't even run test? why do you think there is any problem with RAM? if he removes how to know that it is working without running test?

[_I_]

ram test does not only test ram, it it has other instability problems it could still fail ram test or crash in windows/games

intel does not guarantee it will work with ram faster than ddr5 5600
often it can, but 4 dimms will limit its max speed

basic troubleshooting would be to run it at default speeds and test, then oc one part at a time to see where it fails
but the cpus imc is the only thing overclocked with faster ram, unless he also did any gpu oc