Since I'm John Papadopoulos from DSOG, this was as far as the cable could go. I always make sure to push the cable as much as possible whenever I plug a GPU into the PC. Still, it's clear there is a gap, and this could very well be the reason for the total meltdown. So, it was either an issue that existed with the review GPU sample or, for whatever reason, the cable got loose.
Here are the key takeaways:
a) This can easily happen to anyone. We, reviewers, constantly swap GPUs whenever a new game comes out. So, even when pushing it as much as possible, there is a chance it was not fully seated.
b) The video was taken after I discovered the melted connector and cable. The PC was stable, BTW. I only discovered it due to the smell. For science, I reinstalled the GPU since I wanted to see whether it would work or not. For over 20 minutes, the GPU was working fine at full load with that "gap", without any signs of smoke. So, it's kind of funny that the cable can work fine with that gap.
c) The fact that there is no safe mechanism for when and for whatever reason the cable is not fully seated is a BIG issue, especially for a card that costs 2K-3K euros. When a cable is not fully seated, the GPU should NOT draw power. This simple mechanism could save the GPU.
d) Our replacement is on the way, and we've arranged to send the faulty GPU back to NVIDIA for further examination. (We had already shared pictures and videos with them before publishing the article—they were fully aware of the situation.) Even if NVIDIA hadn't sent us a new RTX 5090, we could have easily purchased multiple units ourselves, so it's not a major issue for us. As I stated in the article, this isn't a "gotcha, NVIDIA" piece. It happened to us—reviewers who've been testing GPUs for over 15 years—so it made sense to document the experience. Nothing more, nothing less.