Unfortunately, I didn't manage more than 500 MB/s with it and therefore removed it again. A real pity. Test environment: Unraid Parity Check with five HDDs, which runs more than twice as fast without a card or directly via the mainboard SATA ports.

Yes works, just plug in and switch on

I do not know for certain, but "it will probably work just fine" to connect this card via a passive x1→x16 riser. It is not a high-power device: it has a small passive heatsink that just gets warm, not hot, to the touch. PCIeX1 cards are allowed to draw up to 10 Watts from the slot's own power pins (https://en.wikipedia.org/wiki/PCI_Express#Power).

So I can't confirm this as far as the card is concerned. However, I was using Linux. Every few hours could indicate a problem with the power-saving functions of the BIOS. The chipset itself is included in Windows 10, otherwise the system would not have recognised the disc connected there. I would do the following: Check the BIOS and check the WWW to see if there are any known problems with the MB. Is there a BIOS update for the MB? Check the logs in Windows 10. What does the system say? Does a standard error occur? Check the temperature inside the case. The controller is passively cooled, overheating due to bad exhaust air can trigger something like this. Read out the DDR RAM to see if there are any memory errors, using Memtest or similar. Read out the SMART output of the connected hard disks. In case of cable problems, the test does not run, aborts as "failed" without a specific reason (often with "at command .... error"). Then open the case and check the corresponding Sata cables. If possible, get a new good cable (remember, short is better!) with metal clip for best fit. Then replace one cable after the other with the same cable, so that all connected discs can be tested one after the other. If the problem no longer occurs with one of them...well then you probably had a cable problem. In my opinion, you should check the fit of the cables and this procedure in any case. Take this opportunity to check whether the card is still properly seated. A card that is not correctly fitted into the slot can cause such errors with "cold - warm" etc., because it slowly "wiggles out". This sounds banal, but it can be the case. Another probable remedy is to move the card from one slot to the other, as sometimes the problem lies in how the BIOS manages the slots. This is mainboard-specific and should be listed in the manual of the MB. And since it comes to mind, have you checked if the PCI-E slot is too slow for the card and you have a faster one? You can also use PCI-E cards in faster slots, the specification is backwards compatible. Last advice: Call the Devolo hotline after these tests are done, with the results at hand. The controller is under warranty and any device can run into problems. Support is usually a bit flippant but friendly. But can only say something if you confront them with the above results.

I have four 18tb on it and it runs stable under linux.

yes, it comes with 5 short data cables