The academic internet is about to be upgraded by a factor of 80….from a 10 gbs backbone to an 800 gbs one according to Sci-Tech Today. Now 10 gbs per second is pretty impressive–that is the speed of the current, famous “Abiline” research backbone so taking an 80X leap is a little breathtaking–and inspiring. Two points come to mind almost immediately: First, it’s a demonstration of the oft-repeated and oft-doubted thesis that fiber optic installations are “future proof” and, second, that there are exciting ideas out there on how to use big bandwidth; exciting enough that people are dramatically improving their networks to accommodate them.
The idea that fiber optic based networks are future proof is regularly dismissed by folks who aren’t technically inclined. I’ve always been impressed by the fact that this claim is most often originates from the least likely source: engineers. Engineer types are seldom seen touting products, much less touting them with phrases like “future proof” that sound like marketing-speak.
But the engineers are really making a fairly limited, relatively technical claim. They are not saying that network upgrades wont’ be necessary, a claim that some appear to think they are making. Instead they are taking almost the opposite position: that upgrades are inevitable and that installing fiber optics now make the inevitable upgrades–whatever unpredictable form they may take in terms of specific protocols–easy and cheap compared to all alternatives. They are saying that installing the fiber-optic hardware is a one-time affair. You won’t be tearing out Fiber to install some new network in order to accommodate the unforseeable specifics of inevitable upgrades.
This upgrade is a good example of why they think upgrades are not an issue; they can squeeze almost two magnitudes of speed increase out a strand of fiber simply by upgrading the electronics/photonics at each end of the fiber. Metaphorically (the actuality is a little more complex) what is being done is that each fiber that formerly carried 10 gbs over light is having new electronics installed that will use 8 different wavelengths–that don’t interfere with each other–to carry 10 gbs each. Said that way it’s easy to see that Dense Wave Division Mulitplexing (DWDM) can make it possible to subdivide the wavelengths being used to achieve dramatic raw speed improvements coupled. The electronics to achieve that can be imbued with whatever fancy new algorithms emerge to make things more efficient. So a huge amount of new upgrade capacity is possible before you ever has to commit to the expense of running more fiber.
For all practical purposes the fiber that gets pulled is all but future-proof. It’s the electronics that get upgraded.
Most of the Sci Tech article is devoted to how folks will use the new capacity–and that’s where the fun really is. With Lafayette getting a shiny new fiber-optic system that we’ll control it behooves us to think about how we’ll use more capacity than most folks can dream of. That’s real fun.
Here’s the way the Sci Tech author finds his way into the question:
Someday, we might conquer the vast distances of space and visit the stars. But right now, on this planet, we are on the verge of eliminating distance itself. And the vehicle for eliminating distance is the next generation of the medium you are now using: the Internet.
The current Net has little to impede you as you search for information…
But if you want to have a live conversation with someone standing in front of the Eiffel Tower, at night, as if they were on the other side of a clear window — with the tower shimmering in more realistic detail than you can absorb — you have two choices. You can either fly there right now, or you can use a PC hooked into the next-generation Internet.
The article emphasizes “telepresence” applications: surgery, virtual travel, advanced cooperation between scientists, access to rare tools, and other forms of “being there.” It closes with:
Even when physical distance is dissolved by bandwidth, we’ll likely find new reasons to get together in real space. As the New World Symphony’s Shook notes: “We’ll never replace live interaction in the same room with other people.”
But we’re getting pretty close.
Telepresence is hardly the end of it…the potential for grid computing, massive multiplayer gaming, and changing our relationship to television or government is enormous. Accessing and manipulating huge data sets may one day be enable each of us does to compose personalized weather predictions keyed to our locales over the course of the day. Or to “look ahead” and help determine the best route for our daily commute dynamically. Each of us could, for the price of a cheap web connection, be our own broadcasting station. Making movies or shows might well become a home craft. A lot of things could change. By far the biggest limiting factor will be imagination.
Give it some thought: what could you do in a community where everybody had a 100 meg connection?