Date: Wednesday, November 21, 2018
As you might know, one of the two RTX 2080 Ti Founders Edition video cards that we purchased directly from NVIDIA failed on us while using it for gaming last week. While we did not get Battlefield V for free, we did get Space Invaders for free, albeit we would have rather done without that freebie.
NVIDIA owned up to 2080 Ti FE card problems as well. Moving forward, NVIDIA support did an excellent job of getting a replacement 2080 Ti FE out to replace our failed card in an expeditious and efficient manner, even if FedEx dropped the ball on the shipment.
We got our replacement 2080 Ti FE on Monday of this week, and got it onto the test bench late Monday night to begin some stability testing. While grabbing some rest, we left the new card running the Heaven benchmark in a loop. Early Tuesday morning, about 8 hours later, we woke to find the 2080 Ti FE chugging along happily. This is of course a great result and were happy with it, which meant we could get to some Samsung VRAM temperature testing later on. Tuesday was a bit busy with family coming into town for the holidays, and having to write an in-depth technical office chair review, so we did not get to the VRAM temperature testing we wanted. However, the kids did want to play some VR games on Tuesday night, so I set up the test system with the 2080 Ti FE for them as it was still ready to go on the test bench. For whatever reason, the kids never got around to the VR gaming, and I went in to find the system sitting on the Windows desktop like I left it, and turned it off for the night and retired.
This morning, Wednesday, we hit the ground running and wanted to see if our new 2080 Ti FE with Samsung VRAM had any temperature differences from the previous Micron VRAM card. We turned the test system on, and were not greeted with a POST screen. Well, that is strange.
This is a new ASUS ROG Maximus XI Formula motherboard has not been through our full testing gamut, so we were not sure of the exact issue. We started by moving the 2080 Ti FE around to the other available PCIe slot, and still no POST. We switched in my Micron VRAM 2080 Ti FE, and the system POSTed just fine using that card in both PCIe slots. We then tried our luck with a GTX 1080 Ti, and had no issues with it either.
Putting our replacement 2080 Ti FE back we were met with the same no-POST, no-signal issue, and we did try multiple times with reseating the card in different PCIe slots. The card's fans do spin up and the logo lights up as it should. The card also generates a good bit of heat on the backplate around the GPU.
We reached out, once again, to NVIDIA's support chat and explained our issue. Once again, NVIDIA chat support was very helpful and has initiated another RMA for us. You can see the quick transcript below.
We have now owned three NVIDIA RTX 2080 Ti Founders Edition video cards. Two of those three have failed on us. While we do "abuse" a good bit of hardware around here, we do not feel that we abused these 2080 Ti FE cards in any way. This replacement card was never overclocked, or used in 3D except in our initial Heaven stability test as explained above. It seems that NVIDIA has not weeded all the way through its "test escapes" just yet. We will let you know how it goes with 2080 Ti FE #4.
It seems that not signing NVIDIA's 5 year blanket NDA has certainly given us a true enthusiast NVIDIA experience, just like our readers should expect.
If you like our content, please support HardOCP on Patreon.