I'll give that a whirl! I've already reset the BIOS, however, I never tried uninstalling the root port from windows. Hopefully it comes up under the device manager, but if not I assume it's definitely toast. I'll try when I get back on Tuesday and let you know!
AMD is known for having the most toxic community on the Internet, they are paranoid beyond good and evil.
They cant handle if their loved company did something wrong and dont get me wrong, the rx480 is a nice card, a costum card would solve this problem, but the reference card (never buy them by the way regardless its Nvidia or AMD ) has issues.
Maybe they fix them with a new driver, maybe not, we will see, but meanwhile those jerks are not able to accept the truth.
RX 480 broke your PCIe Slot, period, you should wait for the new driver AMD is working on and try it again, hopefully they dont slow down the performance.
Anyway, sorry for your loss, AMD does make good gfxcards, hopefully you can have fun with your card soon.
An 80W or so load could absolutely cause an issue in a PCI-e slot if there is any dirt or corrosion. Power protection circuitry would NOT kick in unless it saw a significant load beyond its safety margin. It would appear that the AsRock 970 board has too high of a margin set (or, worse, no protection) - it can probably handle 100W+ on the slot without any dust on the pins.
There's another who claims 3x RX480s killed his board while mining, but the damage is on the +3.3V headers on the 24-pin ATX connector, which disproves that claim. This claim looks more like a prolonged overcurrent situation in the slot - exactly what we'd expect to happen if power protection didn't kick in.
The problem here seems extremely obvious to me, but I've spent 20+ years diagnosing hardware problems.
It is an unknown combination of the following:
1. A long-term high load was incurred on the PCI-e +12V headers resulting in overheating
2. Damage/corrosion caused the Pins 3b and 4b to partially short when new card installed
3. Debris, bent pins, or other factors prevented good contact with the A-side supplies, overheating B-side supplies
4. Motherboard did not deliver equal power to power pins in slot
5. RX 480 did not pull power evenly from pins in slot - favoring pin 3b
6. Pin 4b partially shorted to pin 5b, damaging PCI-e clock signal on the slot and melting the slot
Most of the factors are not related to the video card, but it is the most likely scenario given the overclock and its known, proven, overdraw tendencies. The 150W draw of the card is evenly split between the two power sources (slot and plug), which is generally fine without overclocking. Overclocking the reference RX 480 should be considered off-limits unless AMD can change where extra current is sourced in this condition.
Unfortunately, this is a situation that cannot be fixed by a driver update or even a new BIOS flash update. It's systemic in the circuit topology of the card. The only way it can be done via BIOS is if the card is throttled back, lowering the base clock rate and the voltage. This would effectively castrate the cards performance.
I think most motherboard monitoring would go into a shut down if overall PCI draw is too high. I do not think they monitor individual PCI slots. And as far as the comment about "crappy motherboards" why should that make a difference or even be an issue? It is the PCI card that is at issue, possibly being not in compliance. This is why PCI standards compliance exists. And this mobo in the OP is an enthusiast board. What about OEM mobo's in pre-built comuters like Dell and such? Do people who buy cheap computers, then add a cheap GPU card like the RX480 deserve the potential of burnt connectors, simply because a manufacturer chose to ignore PCI-SIG and it s standards. That's just negligent. This whole thing is a huge mess.
I hope something works for you roquen22. It is a long shot and if not for personally knowing how buggy an asrock can be I wouldn't even recommend those steps. My computer just does that if I happen to unplug one of the display cables from my 970 during bootup and all of a sudden my gpu is no longer recognized. It was a tip I found online that helped a few people with what appeared to be a dead pcie after installing the 960 after release that worked for me. Just clearing the bios is what gets me back up and running so I am not sure what help the rest may be but it was included in what was mentioned I should do at the time. I do apologize if I may have come off rude or sounded like I was insulting your build. I just wanted to get across the issue could have been a combination of things that many seem to agree with now and that amount of power draw is not something unique to the rx 480.
The only way they will fix it is reducing voltage available to the card, which in turn will hinder performance.. The card was bad enough as it was, this will put it on the low end of the modern market at least in reference design... The aftermarket designs might be a different story..
What I dont like seeing already are the reports of screen flickering with the 480, this is an old issue with the fury/furyx and they've yet to fix that issue and its been well over 8 months.
The cards are getting overvolted.
If you go see any number of the links they discuss it. Do a search on youtube, oc3d, jayztwocents, any number of those youtube reviewers have explained it and amd has acknowledged the issue and are working on it...
AMD said they'll have an announcement on it July 5th..