Bandwidth Futures
As a person assigned responsibility for developing the strategic vision of the University’s access to the world’s digital resources, I read with interest yesterday’s report in the Chronicle of Higher Education of an industry-sponsored study about overall inability of available bandwidth to meet demand in coming years. The Chronicle cites concerns from some readers of the Nemertes Research study (you have to create an account to view it) that it is really a subtle strategy to bolster the campaign of the major sponsoring companies for relief from government regulation.
Anyone keeping track of these issues at colleges and universities is likely to have a couple of other thoughts. First, the companies involved have poor histories of recognizing the shape of the Internet-related market with any effective foresight. It is likely that the companies are promoting this study for their own political purposes with respect to deregulation, but that doesn’t make the study’s revelations less interesting or important, in my view.
Second, our colleges and universities are microcosms that demonstrate the ultimate point. Few have sufficient resources to build to anticipated demand (even if we are smart enough to see it coming) — because of the explosive growth of demand, nearly all of us are simply reacting to the pressures of demands already present. Our motives and “drivers” may be different, but the results are the same: demand outstripping supply with disturbing regularity. We are all victims of the accuracy of the predictions from years ago by networking futurists about the convergence of data, video, telephony and other flows of information. If we expect to tap those flows effectively in the future, we’ll have to a much better job than we have to date of investing in capacity-building ahead of demand.
Great post, Chip.
I think one of the ways to approach bandwidth issues more strategically would be harnessing technology that allows people to share the larger files in a more efficient, distributed manner. One possibility is P2P clients like bitTorrent, but with all the fear and paranoia surrounding this technology and filesharing it would be a Herculean task for a university to risk in the current climate.
Nonetheless, if we are really running into serious bandwidth issues, it might behoove campuses to start experimenting with these possibilities, much like Harvard is with their own/homegrown bitTorrent client called Tribler.
Interesting. At Educause, I had the opportunity to talk to a friend who has been heavily involved in I2 and NLR. His question for me was, given all the bandwidth that those projects can (and in many cases already are) affording member institutions, what the heck should they be doing with it (from a teaching and learning perspective). I was a bit taken aback, thinking how a small school like UMW is just struggling to meet the increasing, legitimate bandwidth needs of our current students and faculty.
What’s your perspective on that dichotomy? Are we seeing a situation of haves and have-nots, with large R1 and R2s (who can afford the hefty membership fees for those consortiums and networks) facing an actual *excess* of bandwidth while smaller (particularly public) schools struggle to figure out how to simply afford larger commodity connections?
A couple of quick thoughts for Jim and Martha:
For Jim, I agree — we need to be exploring every means possible to make efficient use of bandwidth, even if the “precedent use” for a particular technology makes institutions a little nervous. That said, I’m a bit skeptical about the capacity of any technology tool to have a sufficient impact — this problem, it seems to me, calls out for a re-establishment of the economy of distribution of bandwidth. When I figure out what I mean by that, I’ll post more thoughts here.
For Martha — yours is a related point to the above. Having been around for some of the early days of this discussion, I understand the original notion that the highest capacities of networking would be required for big-university research. From that notion, it seems natural that the research institutions would be the architects, builders and primary beneficiaries of such consortia. But the world changed while that was going on. My guess is that the proportion of bandwidth consumed at each of those institutions today for traditional research activity is smaller than that used by the institution for other purposes. To the R1s’ credit, I think most would be happy to invite the non-R1s to the party by attractive pricing of connectivity when you reach a node of their national-level networks. The problem for the rest of us is the last mile (for us, the last 50 or so miles; for others, much more). CIOs in similar places to us are burning a bunch of mental energy trying to figure out that part. I have a feeling that both government and the R1s have a role to play in that part of the problem related to their “public service” missions, but I don’t have a clear picture of that role yet.