Those switches can provide power over 4 or 8 pins (configurable).
Passive POE is what all Ubquiti ( www.ubnt.com
) products run off of, and as a result the entire outdoor wireless industry products do as well.
In addition to that, many IP camera and phones are compatible with this method, "passive POE". In addition IEEE devices work fine when plugged in and 48V is selected for 4 pins, respecting a 36 watt / .75A limit. The same .75A (18 watt) limit is enforced when set to 24V 4pin.
Vendors in outdoor wireless (like Ubiquiti) have gone to all 8 pins being powered to accommodate higher power devices.
And with all 8 pins being powered at 24V a 36watt load can be accommodated.
This perfectly suits an Intel NUC (motherboard power header accepts up to 24V DC) but requires a separate large, power injector/extractor (as does your product too, apparently) and cutting the opening of the NUC to access the motherboard power header. This outdoor wireless industry has equipment running in small enclosures at the bottom of communication towers, water towers, grain elevators, etc. and space is a premium. In addition to the space constraint, the the wiring mess these external power extractors cause also makes them undesirable, but currently required.
Netonix markets these switches specifically to this industry to help eliminate power injectors being needed in these small enclosures and provides a solution to power and manage all the necessary equipment directly from the switch.
This works wonderfully with routers, cameras, etc but there is currently no way to directly have an Intel/AMD X86/X64 based computer plugged directly into these switches as a small headless server for network monitoring(SNMP), HTTP servers, etc. I found your product and it's modular approach and was hopeful that it could provide the solution (there is apparently no POE product/module).
A system with 2+ 10/100/1000 copper ports, one of them supporting POE input power would be very useful (or one of the non POE ports could also be SFP, but SFP is not required especially at a large cost impact).
A 48V input on 4 pins would work for both passive POE and IEEE POE or just accepting 24V on all 8 pins would probably be cheaper as there is no step down of the voltage required (as in the Intel NUC example sited above). Either would work fine in my scenario. I'm sure such a device would be very useful to the outdoor wireless / WISP industry, and I'm sure other as well, as there are currently no reasonable, affordable small PCs that run off POE.
Quite the untapped niche....