Doesn't the addition of edge bandwidth just complicate that problem?
The real issue for intra-dorm bandwidth depends upon the size of the trunks between the dorm and the layout of the network. Just giving every individual the ability to pull more bandwidth doesn't help.
The truly interesting thing to see here is the network infrastructure that supports the bandwidth.
It is hard to justify such a significant investment on edge connectivity unless all of the core network functions can handle the load. For 1Gbps bandwidth to actually be useful, all of this bandwidth must be aggregated back somewhere and all of the other shared network functions must also be geared for this incredible bandwidth.
For example, the connection from the campus to the Internet is a bottleneck. If the campus has 45 Mbps to the Internet and this must be shared by everyone, then, as a practical matter, no one will get more then his/her share of the 45 Mbps no matter how much bandwidth his/her desktop is capable of handling.
To take advantage of the speed, you need to invest in a backbone capable of handling all of this bandwidth (so that it is not oversubscribed and therefore irrelevant), sufficient internet capacity, large servers (with 10 Gbps NICs) and a lot of other infrastructure.
Additionally, few laptops will be fast enough to take advantage of the Gbps speed. I doubt whether a PCMCIA adapter can support the throughput required.
The ability to use Voice over IP and streaming video is nice, but can be done over a 100 Mbps pipe. Additionally, the "coolness" factor will probably serve the University's marketing well. Additionally, it probably means that the University will have more years before required upgrades. However, I do not see the justification for this.
Slashdot Mirror
Doesn't the addition of edge bandwidth just complicate that problem?
The real issue for intra-dorm bandwidth depends upon the size of the trunks between the dorm and the layout of the network. Just giving every individual the ability to pull more bandwidth doesn't help.
The truly interesting thing to see here is the network infrastructure that supports the bandwidth.
It is hard to justify such a significant investment on edge connectivity unless all of the core network functions can handle the load. For 1Gbps bandwidth to actually be useful, all of this bandwidth must be aggregated back somewhere and all of the other shared network functions must also be geared for this incredible bandwidth.
For example, the connection from the campus to the Internet is a bottleneck. If the campus has 45 Mbps to the Internet and this must be shared by everyone, then, as a practical matter, no one will get more then his/her share of the 45 Mbps no matter how much bandwidth his/her desktop is capable of handling.
To take advantage of the speed, you need to invest in a backbone capable of handling all of this bandwidth (so that it is not oversubscribed and therefore irrelevant), sufficient internet capacity, large servers (with 10 Gbps NICs) and a lot of other infrastructure.
Additionally, few laptops will be fast enough to take advantage of the Gbps speed. I doubt whether a PCMCIA adapter can support the throughput required.
The ability to use Voice over IP and streaming video is nice, but can be done over a 100 Mbps pipe. Additionally, the "coolness" factor will probably serve the University's marketing well. Additionally, it probably means that the University will have more years before required upgrades. However, I do not see the justification for this.