Question Did I ruin my whole computer?

[uno99]

I wanted to upgrade my graphics card to a 6900xt, but first had to replace the PSU. Went from Seasonic 650w to a EVGA 1200w. Being a bit lazy I decided to use the SATA cables from the old PSU, I put everything back together including the new GPU and when I turned it on I smelled burning and was presented with a No boot at startup.

I realized my two SATA drives were burned up but my M.2 and 2 HDDs seemed OK. The m.2 was corrupted or something so I reinstalled windows on it and everything seemed OK until I started playing Cyberpunk. The game crashes very often, although sometimes I can get an hour before a crash. Often every few minutes it crashes.

Suspicious, I tried running Prime95 and I am getting crashes in that too. If I stop a test fast enough I can generate a report:
FATAL ERROR: Rounding was 2.020321878e+10, expected less than 0.4
Hardware failure detected running 960K FFT size, consult stress.txt file.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected running 960K FFT size, consult stress.txt file.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected running 960K FFT size, consult stress.txt file.

What's everyone opinion here? Is everything ruined or could I maybe try replacing a few of the components?

Currently running is Asus B550, 16GB dual channel, AMD 5600x, 1 tb M.2, 6900xt, EVGA 1200w P3, 2 HDDs, although the oldest one which is about 4 years seems to be dying, has bad sectors.

 

[COLGeek]

As you have found out, cables are not universal between makes/models and you may have permanently damaged your system.

I would assume that everything connected to the SATA power cables has been damaged and that includes the motherboard and things attached to it.

Might be time for a new, new system. You are not the first to learn this lesson, nor likely to be the last.

 

[punkncat]

It isn't clear to me, are you using the drives that were hooked to the incorrect cables, again on the proper cables? Or did you replace those?

If you have a HDD that is in process of failing it is best to disconnect it. I would suggest backing up any critical info on it and then perhaps a wipe and dispose. I often put drives I know are failing into an external drive caddy for such process.

Otherwise, it is possible that you damaged other components and in particular the power supply which is a real shame in this situation.
Cheap first method would probably be to consider one of the better "consumer level" power supply testers. There are many on the market and where they are not the end all be all of knowing all that could be wrong, assuredly if it doesn't pass snuff on one of those it's time to replace.

Have you considered breadboarding the system outside the case, use the old PSU with proper cabling and run a test on minimal parts, then old known "good" and then perhaps swap PSU, then add new GPU to see at what point (if not all) you experience these errors?
edit- do not use the old cables that were hooked up incorrectly (SATA) as those are assuredly burnt up.

 

[drivinfast247]

I would start by running memtest86 on your RAM to verify its integrity. Then I would replace your storage device and install Windows. Then you can start troubleshooting what else may be damaged and replace as you go.

Do you have your old GPU to avoid damaging the 6900xt?

Make sure you have replaced ALL old cables.

 

[uno99]

It isn't clear to me, are you using the drives that were hooked to the incorrect cables, again on the proper cables? Or did you replace those?

If you have a HDD that is in process of failing it is best to disconnect it. I would suggest backing up any critical info on it and then perhaps a wipe and dispose. I often put drives I know are failing into an external drive caddy for such process.

Otherwise, it is possible that you damaged other components and in particular the power supply which is a real shame in this situation.
Cheap first method would probably be to consider one of the better "consumer level" power supply testers. There are many on the market and where they are not the end all be all of knowing all that could be wrong, assuredly if it doesn't pass snuff on one of those it's time to replace.

Have you considered breadboarding the system outside the case, use the old PSU with proper cabling and run a test on minimal parts, then old known "good" and then perhaps swap PSU, then add new GPU to see at what point (if not all) you experience these errors?
edit- do not use the old cables that were hooked up incorrectly (SATA) as those are assuredly burnt up.

Yes I am now using the New PSU cables. Fortunately I did use the new ones for the MOBO and GPU.
I've already got a backup of important files from the dying HDD. I'm not sure if it was dying before this horrible mistake or not.

The old GPU has been sold off so I can't test with the old one. And the new GPU certainly doesn't work with the old 650w.
Maybe I should just RMA the PSU since I just bought it new. Unfortunately the GPU was purchased new for $900 CAD. I REALLY hope that is not part of my problem!

 

[uno99]

I would start by running memtest86 on your RAM to verify its integrity. Then I would replace your storage device and install Windows. Then you can start troubleshooting what else may be damaged and replace as you go.

Do you have your old GPU to avoid damaging the 6900xt?

Make sure you have replaced ALL old cables.

I sold off my old GPU, and the new one was bought used so I can't RMA it if there's a problem. REALLY hoping that it's gonna be OK there.

I'll try a memtest.

 

[uno99]

Ok so the Memtest was a pass. I hope that's a good sign.

Would it be a good idea to just try replacing the m.2 drive? Is it:
1 possible to fix my problem, and 2 would it be dangerous to the new m.2 if my motherboard is the fault?

 

[COLGeek]

Try simply re-installing Windows before replacing.

 

[John Chesterfield]

Maybe I should just RMA the PSU since I just bought it new. Unfortunately the GPU was purchased new for $900 CAD. I REALLY hope that is not part of my problem!

Your problem is that you used the old cables and as said above, these are not universal and you may have damaged your hardware beyond repair.

Unfortunate, but true. A hard lesson indeed.

 

[uno99]

Your problem is that you used the old cables and as said above, these are not universal and you may have damaged your hardware beyond repair.

Unfortunate, but true. A hard lesson indeed.

Yes thanks, I know that's what caused the problem. Looking for a solution to my current situation and trying to avoid scrapping the whole rig.

 

[John Chesterfield]

Yes thanks, I know that's what caused the problem. Looking for a solution to my current situation and trying to avoid scrapping the whole rig.

The only solution is to get known good parts and replace them as stated above until the system is working as it should, ensuring you purchase from somewhere with a good returns policy. Because you'll need that. I'm sorry I can't give you a more specific answer, but you've potentially damaged several parts so it's a case of replacing stuff until it works properly. I can't possibly say this is definitely the issue or that is definitely the issue, how can I? There are several links in the chain that could potentially be damaged, so if you're looking for me to pin that down to one, you're in for a long wait. You could be lucky and only one part is damaged, but the above is the only way to ascertain which components are actually damaged.

That's why I said harsh lesson. I didn't say it to gloat, I knew the reality of the situation you are in and me telling you anything other than the above would be doing you a disservice.

 

[uno99]

Problem solved! It was just an out of date BIOS?!
So without known good parts to swap things in and out, I swallowed my pride and took my rig to a computer store. They told me it was just the bios. I'm a bit shocked since I didn't think my bios was that old, but sure enough everything is working great now.