• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

7950X3D - Weird ram behavior

Soldato
Joined
13 Jun 2009
Posts
4,239
Location
My own head
Putting this thread up to capture anyone else getting RAM issues with the 7950/7900X3D chips.

I'm running the following:
  • 7950X3D
  • Asus B650E-E (Running 1222 bios)
  • 32GB Gskill 6000 CL30 ram
  • Sapphire 7900XTX Nitro+
  • Corsair 1000W RMX
After installing the rig and booting following a bios flash, I experienced a C5 error. I purchased the RAM from someone here, so I know it was working perfectly fine in their system and naturally dived into the fact that the RAM might have an issue as everything was new.

I'm able to run the system on a single stick in A2 or B2, but one of the sticks will run in neither slot, which makes me think it could be a RAM issue. However after seeing this video (https://www.youtube.com/watch?v=M5rkxxqzKRk) and seeing that they used 4 different DDR5 kits I'm wondering if my conclusions could be wrong.

I've not try reseating the CPU because the PC works perfectly fine on the other stick in any position, but curious on thoughts if that could still be a factor (I'd expect weird behaviour with either stick with a cpu mount issue?)

Is anyone else seeing C5 ram issues with these new chips?
 
Last edited:
I've had 'issues' every now and again, same Mobo/CPU/RAM.

Just now for example, went to run the latest Ryzen Master sofware (release today and now supports overclocking on the 7950X3D), it wanted to reboot to set something in the BIOS and BAM... wouldn't POST and stuck at C5.. 3 power off/on cycles by me and still C5..

So I did a CMOS Clear (turn off PSU switch, hold Clear CMOS for 5 seconds, then apply power again and boot), it trained and straight in to the BIOS.

I had this when messing with MCLK settings and had enabled 'Memory Context Restore' which I then had to CMOS Clear to reset.

Touch wood, after stopping messing and having done a CMOS Clear, it seems to be fairly stable (no issues playing games for about 8 hours so far).

I wondered if my board might be faulty after I watched that video this morning and also wondered about my CPU.. Only if a full CMOS clear didn't work (remember to ensure the PSU is powered off, the motherboard RGB needs to be 'off' before pressing the Clear CMOS button) would I bother doing that, as I had already checked it was seated correctly when installing..

My other thought is the AIO Cooler (EK CR360) just says to tighten the retaining screws, it does not say to only do these finger tight or not to us a tool, I presume they are mechanically designed to bottom out and have springs that will apply the last bit of pressure..
Strange.

I've just had my ram RMA approved so see if that fixes the issue, if not then I'll go through reseat and AIO fitting and see if that resolves it before looking at mobo RMA (how Jay solved it).

Like you, with my Deepcool LT520 there is no guidance on tightening the cooler. It has thumbscrews but they have phillips heads so tightened them up past finger tight, can't find any guidance on their website.
 
It does sound like your RAM.. but it does make you paranoid! Did you try the one stick that did work at EXPO settings (6000MT/s)? If that worked, it has to be the RAM or it's possessed.. ;)

I did double check the AIO and instructions and I was correct, it uses the words 'fully tighten' and indeed the entire purpose of the spring is to provide uniform force on each corner when bottomed out.. I did loosen it slightly as a trial, but absolutely no change what-so-ever in temps which is mid 30's in the BIOS (windows has crept up at idle with HW Info etc running to 45-48 most of the time for Tctl, but the CCDs themselves are quite cool and the ASUS socket temp probe is still in the 30's..)

I've been doing more CO Undervolting tests last night and actually the only time I've had C5 is Ryzen Master, when it reboots to do reset / change something in the BIOS it always C5's. However, nothing I set EXPO I / EXPO II myself ever caused a C5, just retraining, it's seemingly quite stable, the only crashes have been when doing CO offsets < -30 on some cores but nothing that looks memory related at all.

I can't say I'm ecstatic about the impact on boot of memory training, but when it's dialled in it's just something to get used to..
Not tried EXPO actually was waiting for 2 sticks but might give it a whirl.

Definitely, a kick in the face when you've spent 3k, but also expected when we're operating at the edges of desktop right now.
 
Back
Top Bottom