It depends on how you want to implement it. There are two ways in which you can implement it, using packet based (real time streaming) and the other one is progressive(based on http, which requires no streaming server like darwin). Both has its own pros and cons.
And for the processing capability it depends on the resolution of the video you are capturing and want to transmit it. If you want a option for the users to select different resolutions liek youtube(like 360p, 480p) then your server has to have the capability to do encoding of different formats, so the double horse power is required. And the encoding server has to be very powerful because it handles encoding as well as transferring the data.
So selecting a server depends on the input & output formats(implementation, file format, codec, resolution), i.e the format IP camera captures and the format which you let your users choose.