Tyranny of Numbers Two: Why cellular carriers can't meet data capacity
As I passed through the gates of the Mobile World Congress in Barcelona, Spain last week, it struck me that the protests outside served as a metaphor for a collection of conference keynotes that warned of the impending wireless capacity crunch. Neither had much to do with the sea of booths showcasing creative new data-intensive apps and ever-more capable smartphones and tablets. But sooner or later, both will prove to be disruptive to those who are trying to make a living selling the products on the show floor.
It only took a few days for the demonstrations along the Plaza de Espaňa to chip away at the exhibitors’ bottom line. The protests -- at first by public transport workers and increasingly by students objecting to budget cuts -- swelled from a curiosity into an impediment as they grew larger and more forceful. Traffic sputtered after police blocked the main entrance and forced attendees to exit out the back. Some left early. Others arrived later the following day to avoid the crush.
To be sure, the capacity crunch will take longer than a few days to cut into the mobile world’s bottom line. But it will. And unless the industry re-focuses its attention and creativity from the exciting handheld advances and onto the more pedestrian matter of ensuring wireless capacity, the cuts could be deep.
LTE, the consensus 4G solution, won’t solve the problem. By build-out, it won’t even keep things from getting worse than they are today, when capacity is already so strained that AT&T and Verizon are alienating long-time customers by curbing bandwidth bloat with limits on “unlimited” plans. (Do carriers commit to anything when we sign those two-year contracts? Or just us?)
So how bad is it going to get? Hang with me, and I’ll take you there.
Let’s start with this tidbit, from the keynote address of Franco Bernabe, Telecom Italia’s Chairman and CEO: Smartphones represent 12 percent of the mobile device volume today, but 80 percent of overall data demand. From that, we can infer that the average smartphone consumes about 30 times more wireless bandwidth than the typical feature phone.
The disparity doubtless will widen as new apps drive new and varied reasons to draw content from the web. A 2011 Nielsen study, for example, reported that the average American smartphone user consumed 435MB per month -- 89 percent more than the 230MB a month they consumed a year earlier. Couple that with the growth trajectory of smartphone shipments -- Gartner says they grew 59.1 percent globally in 2011, to 472 million -- versus that of feature-phone sales, which are flat and forecasted to dip slightly over time.
With that as a backdrop, it’s not hard to see why Cisco is forecasting that mobile data consumption per user will mushroom in 2016 to 18 times what it is today. Combine that with Gartner’s mobile device forecast (he writes and he multiplies!) and you end up with overall data demand in 2016 of more than 26 times what it is today.
So that’s the demand side.
On the supply side, LTE promises improvements in peak data rates of 30x or more over the same time period. But that’s just for an individual connection. In the aggregate, capacity improvements will be far less. A study commissioned by Ofcom, which oversees the UK’s airwaves, forecasts improvements in spectrum efficiency in the 3x range by 2016.
Data demand growth of 26x, but capacity growth of only 3x? Ouch.
Call it the Tyranny of Numbers Two.
The original Tyranny of Numbers was the name given to the wall that the microelectronics industry was hurtling itself toward in the late 1950s. Everyone understood the math. Like the wireless industry today, the applications were progressing far faster than the plumbing was capable of supporting. Programmers contemplated ever-more complex routines that could solve ever-greater problems. But what good were these sophisticated operations if they required more real estate than the city of Barcelona and more solder than a trade school?
It was a problem solved -- independently and at roughly the same time, so the story goes -- by Bob Noyce, then of Fairchild, and TI’s Jack Kilby. The solution: integrated circuits, or multiple, inter-connected electronic elements on a single chip. Like a pilot from movie "Top Gun", the chip pointed the jet skyward just a few yards ahead of the mountainside.
That new trajectory has held to this day. It’s known as Moore’s Law, named after Gordon Moore, the industry pioneer who first graphed the new pace of advance. Moore, who co-founded Intel with Noyce and Andy Grove, declared decades ago that the chip industry was on a path to double the number of transistors it can place on a single chip every 18 months. It still holds true today, more than 30 years later.
Sure beats cordoning off Barcelona to lay out a smartphone, doesn’t it?
Which brings us back to the problem at hand: the Tyranny of Numbers Two. The carriers understand the problem, as they should. But in order to solve it, they’re going to have to tap into the plentiful bandwidth available from wired broadband technologies. That’s something they’ve been reticent to do, because it means allowing their customers’ data and voice traffic to leak outside of their control.
The most promising wireless offload option is WiFi, because it’s already so well deployed, and because it’s already a standard feature in smartphones. The carriers are slowly swallowing their reluctance to adopt WiFi. They’ll need to move faster, though, because there are still some hurdles to overcome. And because data demand continues to grow unfettered.
If they don’t hurry, the Tyranny of Numbers Two will kick them where it hurts the most: in the bottom line.