The list of compatible GPUs for the TS-h1887XU-RP is on the thin side, with a few variations on the GeForce 1030, or the NVIDIA P1000 4G (which I have) or the NVIDIA 620 2G. I’m running QuTS Hero 5.2.7.3297 right now.
Is anyone running anything different in theirs? I’m considering trying the Intel Arc 380 to see if it does more transcodes than the P1000 - reports say it can. Has anyone tried this hardware in theirs?
I’m not sure if unsupported hardware even works. I tried it once with a network card that wasn’t on the list and it didn’t work at all. QNAP may not have drivers for it.
What are you hoping a GPU will give you in QNAP world?
I think I answered my own question, with the answer being no for the Intel ARC GPUs being supported.
The Intel ARC GPUs need Linux Kernel version 6.2 at the minimum, and the kernel QNAP is using on the current QTS/QuTS versions is 5.10.60.
As I stated, I wanted a card that would support more transcodes, if needed. I run a Plex server. Some who are requesting access are seniors who live in my building, and they have older smart TVs that may not always support the encoding format I’m using, so I want to be able to cut down on stuttering that might happen due to multiple transcodes occurring by putting in a card that can support more transcodes at a time. Plus the ARC series supports AV1 codecs.
Ah. Makes sense. Plex will do transcoding w/o a GPU but if you have a bunch of people trying to watch something that needs transcoding, it would probably bog down…
i think the intel arc card need some specific requirements to work (resizable bar) and im sure drivers are not available in qnap OS. I would see if you can fit a nvidia card in your unit. you can try to find a card that fits the size and see if it works. I know the nvidia drivers in quts hero OS support some rtx 30/40 series cards.
I already have the NVIDIA P1000 in the device. In checking other sites, I’ve seen that the Resizable BAR isn’t required for the ARC series to work. I guess I’ll look at other devices or possibly something else in the NVIDIA line.