It was posted here two and a half years ago.

There was just an article earlier today (Orange Goo) about a material that helps absorb shock, so why not line the inside of the helmets with the stuff?

"There is no great genius without a mixture of madness" - Aristotle (384 BC - 322 BC)

We still don't have a cure for it, so you are just telling the person at present 'Hey, you have Alzheimer's! Good luck with that!". I know, there might be treatments that can take advantage of early detection some time in the future, but at present it would just be pretty depressing.

...but the servers aren't the only thing in a data center. If the switches and routers can't take the higher heat, then you aren't going to get much use out of those servers.

What is the carbon footprint for the manufacture of this item? How long does it have to be run before the amount of carbon that went into its manufacture is balanced by the amount of carbon not being released into the atmosphere?

I really don't know where you are getting your math from.

7 participants should have a total of 28 UDP streams (2 video and 2 audio each) and two TCP streams each (for H.245). Add another two streams each if there is H.239, but H.239 works within the bandwidth envelope of the conference itself, so no added bandwidth. So I don't see how a 7 participant conference would have more than 42 unicast connections, 56 with H.239.

If the video streams are 140K and the audio is 56K you are looking at around 1.4M bandwidth before overhead. As most of that is UDP, there isn't that much overhead as compared to TCP streams.

And no, the 45 participants were all live, active participants. The MCU was in a data center, but the endpoints were all over the place, some over LAN links, some over DSL. And they were using H.239.