Hi  @lukass2000, the standby power consumption is show at datasheet, you may check at SG2428LP(UN)_V1_Datasheet | Omada Network Support, SG2428P(UN)_V5.30_Datasheet | Omada Network Support. SG2428LP use 8.9 Watts in Standby, SG2428P use 14.4 Watts in Standby, the difference is the the use different power supply module, standby power of SG2428P power module is higher. Also SG2428P have Fans also use power.

Anyway, back to the demand, if your friend don't need PoE output currently, but will need future, I think SG2428LP is more suitable, it is fanless and lower poe budget/ lower power consumption. It is good choice.

  @Penry.huang 
 

I spoke to him again, he will now install two EAP650s and power them via PoE.
The SG2428P would therefore only have the advantage that all ports could be supplied with PoE, or?

Isn't it somehow more efficient when PoE devices are also used, so that the SG2428P has any particular advantage?

Thanks :)

Hi  @lukass2000, yes, all port of SG2428P support PoE+ output, and the total PoE output power is higher.

Definitely, SG2428P is a good choice, it is suitable for scenarios the have a lot of powered device like EAP, or don't have so much powered devices yet but want future expansion.

  @Penry.huang 
 

The SG2428P is almost identical in price to the SG2428LP.

Therefore, the idea is to just get the SG2428P, even though the SG2428LP certainly has enough PoE ports right now.

But more and more PoE devices are becoming available.
I've read that managed switches vary quite a bit in energy efficiency.
Hence my question about performance, but perhaps I haven't thought it through completely 
Let's assume I connect the two EAP650 to the switch via PoE (and would fully utilize the Wi-Fi network).
Can we then say that the SG2428P uses 15.6 watts, two EAP650s use a maximum of 13.5 watts, resulting in a total of 42.6 watts?
The SG2428LP would use 8.9 watts, two EAP650s use a maximum of 13.5 watts, resulting in a total of 35.9 watts.
Or is it all dynamic and therefore no comparisons can be made?
Thanks :)

Hi  @lukass2000, I think your calculation principle is total correct, but the power consumption of switches and access points does indeed vary dynamically.