Comments on: Virtual reality to get its own network?

Group says Neuronet, separate from Internet, to be launched in 2007 purely for virtual-reality games and business apps.

[solrosenberg]

Doomed from the start

Reinventing the Internet is not quite reinventing the wheel, but it's close.

[ThePenguin]

Doomed?

I dont see this as reinventing the internet.

Why would this be doomed? If you have huge bandwidth requirements as well as a need for low latency, this service would be ideal.

Much like the Internet2.

[punterjoe]

reinventing the wheel or refining a spin-off?

I have no problem with a dedicated network for a specific purpose. It's like FTD vs UPS. Of course it's much harder (& more expensive) to create & implement a unique network from scratch, so I fail to see their business model. At least the internet would let someone build on an existing infrastructure - even if it's not ideally suited to the task.

[Dachi]

Agreed

Internet2 and NLR are sort of a testbed for new network technologies already, but they still use routing products from companies like Cisco and lease fiber from companies like Quest and Level3.

The vendors support Internet2 and NLR because they get a testbed/lab for new technologies in development for the internet (like IPv6 and more secured replacements for DNS, SMTP etc.).

The vendor backing and government/university funding are the reason those networks are able to stay alive.

I fail to see the point of starting from scratch with little to no vendor backing to create a network with only a specific finction in mind.

Everyone else is setting out to make a platform to move any data quickly, and they seem to be doing fine so far.

With todays "internet" technologies you could get dedicated fiber and a couple backbone routers that will put 640 gig of traffic from point A to point B anywhere in the world at nearly the speed of light.

The only reason people on the internet peer off traffic and wish it the best instead of running all their own dedicated fiber is is becasue it is orders of magnitude cheaper, but there is no law saying they have to.

I fail to see how building a network from the ground up to push virtual reality data is going to make fiber optics any cheaper or more efficient.

The whole "virtual reality" and making networks faster are really Apples and oranges anyway.

What do they intend to do? Replace fiber optics with something faster? Compete with Cisco and Juniper and build a router that will move data faster than wire speed and handle more than ~640 Gig a second?

Since replacing the data transport used for the Internet makes such little sense, the best I can make of their plans are that they want to replace the TCP/IP protocol stack with something lighter, but with a more limited function at the expense of having to build an infastructure to replace pretty much everything we use today for the sake of saving _maybe_ 10 ms and likely at the sacrafice of reliability for....(drum roll please).... virtual reality!

Uhm yeah, good luck with that one.

[dazirius]

Mentioned here

I got to this story via another piece written by someone else who claims this whole thing could be bogus.
http://blog.rebang.com/?p=1137
Have a look.

[dazirius]

Oops

Should've read the rest of the updates on the other article.

Sorreh, still, it's a nice read anyhow.