What hardware components possibly got damaged from this catastrophe?

[SinanDira]

Hello! About two months ago, my computer broke down out of the blue and would since then give me a blue screen whenever I tried to boot into Windows with my GPU's driver installed. No driver, no problem. However, to demonstrate that the issue might be far more complicated than it sounds, I have to detail to you the chain of events and observations which surround it:

1. In early 2016, about half the pins of my motherboard's PCI-E x16, the one seating my dedicated GPU, got plucked out; the pins literally got severed. (This was a shipping accident, after about 2 years of constant -but safe- air transportation between my school's city and hometown during vacations.) However, the computer and the card kept working normally, with the only difference I observed being that the card would from then on work in x8 mode instead of x16.
   
   
2. For over a year thereafter, everything functioned normally, until about two months ago when the computer froze out of nowhere. I don't remember what I was doing, probably playing some intensive games, probably just browsing. From that moment onward, a blue screen would emerge right after the Windows logo. After some tinkering in safemode, which was working normally, I came to learn that the computer booted into Windows normally had I uninstalled the GPU's driver. If the driver is installed, the issue recurs regardless of the version; I tried older and newer alike.
   
   
3. One thing worth noting is that while examining the computer right after the incident, there was a very intense metallic smell, which I finally traced to the power supply, where a broken piece from a plastic cable tie was apparently jamming the fan. I removed it but nothing changed. Windows kept booting fine without the display driver, and kept giving me a BSD after the logo when the driver was installed. I searched for a fix for days then gave up and put the computer aside for these two past months.
   
   
4. Now, when I got back to it, the computer was stuck in an infinite cycle where it restarts itself after the POST screen. In a final desperate attempt to get things to work, I went to the BIOS and restored all factory settings, but made sure to reactivate the RAID and Intel Rapid Start options before exiting because my system uses an SSD cache. That got the computer to get past the POST without restarting. However, the usual happened; Windows failed to start because of a recent hardware change. The same ♥♥♥♥ing ♥♥♥♥♥♥♥t; one small change in the BIOS and Windows breaks down completely.

It's gonna take me some time before I get Windows back to work. First things first, though, what are the chances of my computer surviving such a catastrophe with little to no damage? The first technician I took the computer to seemed rather rash and unco-operative, but he told me that trying to fix a comp. with a broken PCI-E slot was a waste of his time and that it's likely to have caused damage (to other components?). Before believing anything he says I figured I'd post asking about the odds here! Should I replace any components? Should I attempt a Windows reinstall or is it truthfully a hardware fault?

Specs:
Core i5 3470 at 3.2 GHz
GTX 760 2 GB
8 GB DDR3
Gigabyte GA-Z77-HD3
Corsair 600w powersupply
64GB Intel SSD running as a cache using Intel Rapid Start Technology
WD 1TB Blue system drive
Windows 7 x64 (can't recall if it was SP1 or 2)

Thanks a lot in advance!

[TehSpoopyKitteh]

Unfortunately you are going to need a new GPU and motherboard. Your GPU won't function in any other PCIe slots without those missing connector pins, and those pins are not replaceable.
You cannot simply replace the PCIe slot unless you know specifically which leads you need to solder.

What you need is a motherboard that matches the damaged one's chipset...and a new GPU...

For a compatible parts check I would recommend using PCPartPicker. Just be aware of the form factor user by the case to mount your current motherboard.

https://en.m.wikipedia.org/wiki/Computer_form_factor

[Dr.Shadowds 🐉]

Try this, remove GPU, plug your monitor to your motherboard, and then see what happens when you boot up.

If you want to move your stuff from the SSD to HDD, I recommend you to have a USB 4GB, or bigger, put linux on it, boot from it, then move your stuff over, hopefully you didn't use any encryption software.

What it sounds like is that the motherboard is the issue here, i'm not 100% sure on this, it can also be the PSU as you said you smell something burning from it.

[MeStinkBAD]

You probably pulled out the PCIe pins used for the top 8 lanes. If the first 18 pins are still in place than it should still work. You could try moving the GPU to a different slot and see if that fixes it.

I assume since the computer still turns on that the power supply still functions. It could be the power supply. The only way to know would be to replace it.

If neither of these suggestions fixes it then I suggest trying reinstalling Windows. After that there’s not much else I think you can do.

Good luck.

[TehSpoopyKitteh]

Originally posted by Dr.Shadowds 🐉:

Try this, remove GPU, plug your monitor to your motherboard, and then see what happens when you boot up.

If you want to move your stuff from the SSD to HDD, I recommend you to have a USB 4GB, or bigger, put linux on it, boot from it, then move your stuff over, hopefully you didn't use any encryption software.

What it sounds like is that the motherboard is the issue here, i'm not 100% sure on this, it can also be the PSU as you said you smell something burning from it.

That's usually never a good sign. The connector pins being stuck insuse the PCIe slot could be causing the motherboard issue. Good news is that the damage caused power overdraw (the cause of the burning smell) seems to be localized to the PCIe slot...so at least they can keep the CPU and RAM.

[TehSpoopyKitteh]

Originally posted by MeStinkBAD:

You probably pulled out the PCIe pins used for the top 8 lanes. If the first 18 pins are still in place than it should still work. You could try moving the GPU to a different slot and see if that fixes it.

I assume since the computer still turns on that the power supply still functions. It could be the power supply. The only way to know would be to replace it.

If neither of these suggestions fixes it then I suggest trying reinstalling Windows. After that there’s not much else I think you can do.

Good luck.

It's an extremely slim chance that it will connect properly. Even if it does work, it could cause further damage to the already damaged southbrbride/chipset because the circuit leads are damaged. The issue is similar to those found with overuse of the front loading NES...the slot connector between the cartridge pins and the PCB are connected, but cannot connect properly. Heat buildup will cause the pins jammed into the slot to bend and breaks the leads in the GPU causing further problems.

[SinanDira]

I forgot to mention a somewhat important part. The computer worked perfectly normally for extended periods of time (full time over many days) when I booted using the motherboard's integrated graphics. This supports the notion that any damage or hardware issues are occurring only with the dedicated GPU and/or its damaged PCI-E slot.

Originally posted by Dr.Shadowds 🐉:

Try this, remove GPU, plug your monitor to your motherboard, and then see what happens when you boot up.

Works, as explained above. Does that mean that a new mobo/GPU would solve things?

Originally posted by Dr.Shadowds 🐉:

What it sounds like is that the motherboard is the issue here, i'm not 100% sure on this, it can also be the PSU as you said you smell something burning from it.

Not burning; just metal, like the smell of computer components out of the box but a tad more intense.

Originally posted by 🎃Sir Edmund's Spoopy Kitteh:

Unfortunately you are going to need a new GPU and motherboard. Your GPU won't function in any other PCIe slots without those missing connector pins, and those pins are not replaceable.

Are you saying that all PCI-E slots on a mobo stop working if just one of them gets damaged? Just curious here. As a matter of fact, using the other PCI-E blocks 4 SATA connectors out of 6, and I need at least 4. Talk about genius designs!

Okay, say that the severed pins are the source of the trouble, why would I need a brand new GPU? Wouldn't it do to just mount the old GPU on a new mobo?

[Dr.Shadowds 🐉]

Originally posted by SinanDira:

I forgot to mention a somewhat important part. The computer worked perfectly normally for extended periods of time (full time over many days) when I booted using the motherboard's integrated graphics. This supports the notion that any damage or hardware issues are occurring only with the dedicated GPU and/or its damaged PCI-E slot.

Originally posted by Dr.Shadowds 🐉:

Try this, remove GPU, plug your monitor to your motherboard, and then see what happens when you boot up.

Works, as explained above. Does that mean that a new mobo/GPU would solve things?

Originally posted by Dr.Shadowds 🐉:

What it sounds like is that the motherboard is the issue here, i'm not 100% sure on this, it can also be the PSU as you said you smell something burning from it.

Not burning; just metal, like the smell of computer components out of the box but a tad more intense.

Originally posted by 🎃Sir Edmund's Spoopy Kitteh:

Unfortunately you are going to need a new GPU and motherboard. Your GPU won't function in any other PCIe slots without those missing connector pins, and those pins are not replaceable.

Are you saying that all PCI-E slots on a mobo stop working if just one of them gets damaged? Just curious here. As a matter of fact, using the other PCI-E blocks 4 SATA connectors out of 6, and I need at least 4. Talk about genius designs!

Okay, say that the severed pins are the source of the trouble, why would I need a brand new GPU? Wouldn't it do to just mount the old GPU on a new mobo?

Only way to find out for sure is to replace the motherboard as that's the only way, since you said the motherboard PCI-e is damage, and would most likely be the issue in this case.

[TehSpoopyKitteh]

Originally posted by Dr.Shadowds 🐉:

Originally posted by SinanDira:

I forgot to mention a somewhat important part. The computer worked perfectly normally for extended periods of time (full time over many days) when I booted using the motherboard's integrated graphics. This supports the notion that any damage or hardware issues are occurring only with the dedicated GPU and/or its damaged PCI-E slot .


Works, as explained above. Does that mean that a new mobo/GPU would solve things?


Not burning; just metal, like the smell of computer components out of the box but a tad more intense.


Are you saying that all PCI-E slots on a mobo stop working if just one of them gets damaged? Just curious here. As a matter of fact, using the other PCI-E blocks 4 SATA connectors out of 6, and I need at least 4. Talk about genius designs!

Okay, say that the severed pins are the source of the trouble, why would I need a brand new GPU? Wouldn't it do to just mount the old GPU on a new mobo?

Only way to find out for sure is to replace the motherboard as that's the only way, since you said the motherboard PCI-e is damage, and would most likely be the issue in this case.

It is a precasuionary thing for one. Obviously you do not want to be using integrated graphics given that CPU. Anyway, secondly, what prompts me to suggest a new motherboard AND GPU is that you could have easily damaged the GPU you were using in the first place. The volatage from a 16X PCI-e slot being placed upon only half the connected leads in stead of all 16 lanes can cause heat damage to the existing connector pins on the GPU as well....and judging by how you could still run on all 16 lanes with damaged leads from the slot to the motherboard tells me that some of the pins on the GPU are damaged.

[SinanDira]

Okay, gentlemen, If I did not miss anything, this should be the correct troubleshooting order:

1. Reinstall Windows and display driver.
   
2. Replace motherboard. $110 http://amzn.to/2zR1TFt
   
3. Replace GPU. $210 http://amzn.to/2zSPJfn
   
4. Replace PSU. $50 http://amzn.to/2ie4l1A

What is the likelihood that any other components outside of the above got damaged as well? A piece or two are fine, but I ain't gonna renew an entire 4-year-old PC.

[TehSpoopyKitteh]

Originally posted by SinanDira:

Okay, gentlemen, If I did not miss anything, this should be the correct troubleshooting order:

1. Reinstall Windows and display driver.
   
2. Replace motherboard.
   
3. Replace GPU.
   
4. Replace PSU.

What is the likelihood that any other components outside of the above got damaged as well? A piece or two are fine, but I ain't gonna renew an entire 4-year-old PC.

If almost all else fails, you will only be keeping the CPU and RAM.

[SinanDira]

This is coming along cheaper than I expected. Would anyone suggest a dual setup of older and cheaper cards over the one I mentioned there?

[Dr.Shadowds 🐉]

Originally posted by SinanDira:

This is coming along cheaper than I expected. Would anyone suggest a dual setup of older and cheaper cards over the one I mentioned there?

You mean you want to SLI/Crossfire? Don't as it's not worth wasting your time doing such of thing.

[TehSpoopyKitteh]

Originally posted by Dr.Shadowds 🐉:

Originally posted by SinanDira:

This is coming along cheaper than I expected. Would anyone suggest a dual setup of older and cheaper cards over the one I mentioned there?

You mean you want to SLI/Crossfire? Don't as it's not worth wasting your time doing such of thing.

It's not a bad thing honestly. One video card handles half the frame scan, the other video card handles the other half...Such interleaving allows for better performance....at lower, more efficient clock speeds. But for what the OP needs, you are correct.

[SinanDira]

Okay, moving on, and with a detail that I had forgotten about completely. Once I hit that "restart" button on nVidia's driver installer, the screen displays artifacts starting from the RAID briefing list after the BIOS screen and up until (and including) the BSD after Windows' logo.

I just retested this on a clean Windows install with just one HDD plugged in and the exact same thing happened, except that I could hear the Windows sound after the logo this time around. At that moment, the screen was completely filled with artifacts, and one and a half second later a BSD showed up, again, with artifacts all over it.

I'm going out now to see if I can find any compatible MoBos and affordable GPUs in my town's market. Please update me if you have any better ideas.

EDIT: I just tested the GPU in the other, undamaged PCI-E, and it caused the exact same issue. Is that an indication that the GPU has most likely gone bad?

Would swapping GPUs with my brother's PC be a legitimate method for knowing whether the issue comes from the MoBo, the GPU or both?

[TehSpoopyKitteh]

Originally posted by SinanDira:

Okay, moving on, and with a detail that I had forgotten about completely. Once I hit that "restart" button on nVidia's driver installer, the screen displays artifacts starting from the RAID briefing list after the BIOS screen and up until (and including) the BSD after Windows' logo.

I just retested this on a clean Windows install with just one HDD plugged in and the exact same thing happened, except that I could hear the Windows sound after the logo this time around. At that moment, the screen was completely filled with artifacts, and one and a half second later a BSD showed up, again, with artifacts all over it.

I'm going out now to see if I can find any compatible MoBos and affordable GPUs in my town's market. Please update me if you have any better ideas.

EDIT: I just tested the GPU in the other, undamaged PCI-E, and it caused the exact same issue. Is that an indication that the GPU has most likely gone bad?

Would swapping GPUs with my brother's PC be a legitimate method for knowing whether the issue comes from the MoBo, the GPU or both?

That is quite possibly the case. You used it in a damaged PCIe slot and as you initially described, some of the connector pins from the GPU were still in the PCIe when you removed it from the slot. The Integrated graphics chip on the board is handled through the Northbridge IC and may not be affected by the damages done to the GPU and the Southbridge leads to the PCIe bus slot you used.

So what we are looking at is definitely a motherboard replacement (not entirely expensive for the CPU you and RAM you already have), GPU upgrade, and possibly a new PSU to compensate for the potentially increased power consumption to conpensate for the GPU upgrade. I would recommend a NVIDIA GTX1060 (6GB version), or a NVIDIA GTX 1050Ti if you're on a tight budget. Just make sure your motherboard is compatible with the CPU and RAM you currently have, and the new GPU and PSU, and can fit inside the case you are using. If your broken motherboard is standard ATX in form factor size, make sure the board you replace it with is also ATX in form factor size. Given your CPU, the specific chipset you will need to include is the Intel 70 series (i.e. H77, B75, Z77).

The easiest place to check compatibility for is PCPartPicker's system build option to get an idea for how much it should cost, as well as getting the motherboard and GPU of your choice.