AMD 7700X + 32G GSkill 5600 + Asus B650 Prime = 💩

M

MrWizard6600

Supreme [H]ardness
Joined
Jan 15, 2006
Messages
5,791
Hey guys,

I'm being tortured by what I thought would be a nice upgrade for my girlfriend. I'm in the nightmare hellscape of debugging, so any help/advice is appreciated.

AMD 7700X
ASUS Prime B650 A AX
32GB Gskill DDR5 5600 with nasty cas latencies :(
Sapphire Radeon RX 6650 XT
Inwin (?) 2TB NVME SSD
Seasonic 650W and a Lian-li something-or-other

I did this on release day for B650, so everything came from Microcenter except the mobo, which came from newegg...

and the things an obnoxious lemon.

- First install, the thing wont run at its EXPO (AMD's XMP) profile, so it ran at DDR5 4200 I believe
- apps start crashing regularly, start seeing errors in OCCT memory stress test
- start swapping sticks around, eventually the thing wont boot
- update the bios to a beta published on ASUS page, I believe its v805.
- things seem better,
- eventually they deteriorate, am able to produce more errors in OCCT
- start testing one stick at a time; with the second stick the thing wont POST
- return the ram, get new ram from microcenter (thanks guys!)
- things are stable for a while, passes several 4-hr+ runs of OCCT... but also start deteriorating
- re-install windows, OCCT memory test crashes after 8 or so minutes.

and thats where I am. At this point I'm maybe 50% confident the board is bad. I've never had a bad CPU before but i'm willing to believe that could be the problem. Similarly, the way that the computer works for a while and gradually becomes less stable imples some kind of problem flow im not familiar with. The fact that these problems wont stay still makes me think that maybe its tempature related or theres some heuristic or driver thats fucking with me.

At this point I am no longer in the country and my poor girlfriend has to deal with this garbage. I've told her to throw it into microcenter and I figure I'll pay them to replace the board and put in a new one. I havent done this yet, if you guys have any suggestions I'd really like to hear them.

---

edit: they replaced the motherboard, CPU, ram, power supply, and the SSD: "instability" persists. They also tried moving from windows 11 to windows 10, and they tried a different GPU.

At this point the parts still being used are: the case, the graphics card, and the noctua cooler i put on it.

We're at a complete loss. The guys at microcenter are suggesting I simply buy an intel board and CPU. Its clear we're all completely out of ideas. There must be something wrong procedural, or there is some crippling defect in firmware across all three boards.

---

Regards,

-Geoff
 
Last edited:
If your girlfriend is technically inclined she could reseat the CPU, take some attention on the CPU heatsink, how much pressure being applied.
Only one memory dim would work as a troubleshooting step as well. It could be that one of the slots has some sort of issue.

If not, the route with the guys from Microcenter to help would be a solid choice.
 
yeah so I paid for a new motherboard and $200 (!?) for Microcenter to do a motherboard swap.

The thing ran their 16 hour stress test and just ran for 8 hours today including gaming without any issue.

Now the thing is having trouble staying in windows for more than 5 minutes without a BSOD.

This is exactly the same failure pattern as before: The machine seem stable for a prolonged period of time, then starts crashing very regularly.

I'm at a loss
 
I've seen psu's do this before. The 12v rail regulators were crappin after being loaded for awhile. Check all psu outputs for proper voltages when it starts acting nuts.
 
That was the first thing i tried the first time pendragon1, but it didnt fix the issue the first time.

At this point an OCCT failure with hundreds of memory errors in the first few seconds implies some kind of event has happened that has done damage to the memory hardware, right?

The fact also that the same windows install (read: after changing the mainboard I did not re-install windows), was stable for 16 hours after a stress test implies the drive is fine.

So its not the drive, its not the mobo, its not the ram, I've never heard of a bad CPU (short of a physically destroyed one)...

that really only leaves the PSU.

One other note: the PSU is not grounded. Shes in an old house that doesn't have grounded plugs, so they use "an adapter" (something that simply strips the ground) to plug things in. To my knowledge the only thing that ground is used for is for safety, but im wondering if it might also be some kind of hedge against ESD.

I hate this so much.
 
is the psu new or used? could try a different one. idk know if the ground missing would cause probs, just safety.
 
no its a brand new seasonic Focus 650+, its not the top-of-the-line from seasonic, but it is a seasonic.
 
when they swapped ram is it the same stuff? could try a different brand.
 
  • Like
Reactions: xx0xx
like this
bump.
pendragon1 the ram is exactly the same SKU; they wont want to change skus without some kind of charge right. I'll ask them to give some DDR5 they having laying around --asuming they have some-- a shot.

I updated the issue, at this point they've tried swapping literally every component in the computer: it still "has stability issues".
 
Did u check the motherboard site for the QVL list of supported memory? I only buy RAM on the QVL list as I had run into similar issues when I first built my 6700k system years ago. After multiple trips back and forth it was the RAM that was causing instability and only RAM on the QVL list worker with the MB.
 
You must log in or register to reply here.
Back
Top