One major point of these guides is to save money by purchasing the recommended parts at the time published from the used market.
The point of quicksync specifically is that it quickly outpaces standard CPU transcoding while keeping the cost down. A 50 dollar CPU with quicksync can outperform a several hundred or thousand dollar CPU without quicksync when it comes to plex transcodes.
It is recommended to keep Plex with quicksync as a dedicated system to simplify setup and maintenance. More specifically, this is to avoid complications passing the iGPU through docker or a VM and avoid potential resource sharing issues.
The Xeon you listed does have quicksync, but also costs nearly 3x as much an entire HP 290 or Otis build, and thatās just the CPU.
Since the 290 availability and price fluctuates and is inconsistent lately, I recommend the Otis build. Either of these options should be paired with a NAS, but Otis can also be built out as a multipurpose all in one box if you wish to go that route. Check out the NAS Killer 4.0 guide as well.
If you mean transcodes, with QuickSync that CPU can do 20 easily. 16GB of RAM is completely overkill, most people usually use 8GB. RAM speeds are not important, 2400 or 2666 is fine.
Thereās one G5400T left for $39.99 - no reason to buy new.
After reading Iām seeing that LGA1200 Mobo are about the same in price. But they do not have the LGA 1200 pentiums in stock anywhere. Is a New Celeron a better option for a plex/home storage over the G5400?
Where are you finding a LGA1200 motherboard for $70? Because thats what the 9th gen motherboards are running new at microcenter. The cheapest LGA1200 motherboards I saw there were $150.
The difference between 8th/9th and 10th gen is minimal. Personally I would invest in LGA1151 with an 8th-gen CPU, and when it comes time to upgrade, the 8th/9th gen i3/i5/i7 will be much more affordable than LGA1200. Thereās nothing that LGA1200 brings to the table that LGA1151 doesnāt already have.
Sort of what I was thinking going to go with the 8th-gen pentium then. $40 under budget https://pcpartpicker.com/list/
Anything I should spend a bit more on? Also only 12TB right now as this is first NAS
Your PCPP link doesnāt work - Iād suggest using google sheets and copy pasting the table in here, it works great. PCPP doesnāt help a lot with server hardware.
Interesting the talk about getting 20 h264 transcodes āeasilyā.
I struggle to get 20 1080p h264 transcodes on a brand new Intel Xeon E-2276G, Coffee-Lake gen with Intel P630 graphics. These are 15-20mbps files with DTS 5.1
64GB RAM - NVMe Intel P3600 NVMe
If I try and do any more than 20 it starts buffering - despite disk, network and cpu all being low - which to me suggests the QSV chip is maxing out
I donāt have any experience with the E-2276G. Iād recommend trying to upgrade to 20.04 as recommended in the guide. As far as I know, subtitle transcoding takes place exclusively on the CPU, which is different than video stream transcoding.
Also, if youāre doing anything else on the machine, it changes the environment, negating any of the results stated here.
Note// To remove disk performance from the equation - content was saved on 64GB ramdisk and transcoding also occured in RAM.
RESULTS
QSV
Starts to buffer at around 20 transcodes - with significant IOWAIT - despite files and transcoding stored in RAM. System load is minimal (2-3), and user/system CPU usage is also minimal (20%).
See streams buffering and transcode speeds less than 1.0 for reference.
SUMMARY
Whilst QSV performs adequately for less than 20 transcodes - and CPU usage remains low throughout the test - QSV transcoding generates noticeable IOWAIT despite disk activity being minimal. IOWAIT increases in step with the number of transcodes running. It does not occur once a particular threshold is met- instead it builds slowly as transcodes are added. Once IOWAIT hits approx 60% transcodes begin to buffer.
It is important to remember that IOWAIT isnāt specifically disk related.
The exact same test on NVENC does not see this same IOWAIT side effect, running on the same hardware, same disks, same transcode setup. As a result - 24 simultaneous transcodes run without buffering.
