Hi There!
Know this has been asked a thousand times but things change so frequently with the release of new hardware and software updates.
I’m currently running a Windows 10 Plex Server with Dual Xeon’s but they are older with a slower core speed and I’m noticing stuttering issues when transcoding 4K with HDR Tone Mapping turned on. And transcoding 4K is not worth it without Tone Mapping turned on.
I added a GTX 1660 Super to the mix thinking that would solve my issues not realizing that HDR Tone Mapping was CPU Dependent.
So now I’m back at the drawing board. Since this is a stand alone Plex Server feeding 10-15 users not all at the same time, I’m looking for some Hardware suggestions. I know direct stream or pre-transcoding (Optimizing) the 4K files are the best solutions but I tinker with this kind of stuff, it’s my hobby.
I obviously have a Nvidia GPU so that can be transferred over to a new Mobo / CPU Configuration. And it can be upgraded to a more modern Turing / Ampere card in the future (Even Quadro). With that in mind it comes down the CPU Core Count and CPU Core Speed. I went after Core Count with Dual Xeon’s a few years back. Not thinking Core Speed is just as important. Based everything on Passmark scores and bang for the buck!
I’m thinking that if you don’t have a GPU already, investing in a Intel CPU with QuickSync is your best choice. But since I already have a GPU am I better off looking at a AMD CPU setup over a Intel CPU? I guess I’m wondering if the AMD’s handle the CPU intensive Plex tasks better than the Intel’s knowing that the Video Transcoding is already off loaded to the GPU.
I don’t believe that there is really a quality difference in Video Transcoding between QuickSync & Turning / Ampere Nvidia GPU’s. Could be wrong with that assumption though!
Also curious to know if Base Core Speed is more important than Turbo or Boosted Core speeds in a Plex Server and what core count I should be focusing in on for a system that serves no more than 5 users at a time. Obviously the more cores the better but I assume that you wouldn’t want to drop below a certain base core speed though as that’s my issue right now!! I have 6 Core & 12 Threads x 2 but they run at 2.4Ghz and Boost to 3.2Ghz and it’s apparently not enough for the Tone Mapping to keep up.
Curious to learn others thoughts on the matter.
Thanks In Advance!
