With today’s celebration of World IPv6 Launch 2012, the Internet is doing something new: growing. And that growth comes none too soon as the rise of the Internet of Things places unprecedented new demands on worldwide Internet infrastructure.
Google, Yahoo and Facebook are amongst the leaders in the growing charge to adopt the critically needed Internet addressing system known as IPv6, which is generally reagarded as the viable only solution for the stagnation of the Internet.
World IPv6 Launch 2012 brings together “major Internet service providers (ISPs), home networking equipment manufacturers and web companies around the world… to permanently enable IPv6 for their products and services,” starting today.
Here’s why, by the numbers:
In 1973, when Vint Cerf and his team put together the networking rules for what would become the Internet, they used an addressing system with 32 bits of addressing space – the well-known 192.X.X.X IPv4 system in use today. This gave the fledgling Internet the capacity for 4.3 billion individual addresses; far more than Cerf and his team could even conceive of needing back then.
Obviously, Cerf and everyone else severely underestimated the growth of the Internet and all the various ways it would be used. More than just a system to share files and images, the Internet has become a platform for commerce and communication that eventually dwarfed the telephone network, the only comparable network on the planet.
That growth has led us to the problem we have today. According to Cerf – who took part in a Google Hangout Tuesday afternoon, there are currently 5.5 billion mobile devices in the world. If each one of them were to need an IP address (and that’s likely to be true in the very near future), they alone would require more than the available Internet addresses under IPv4. New devices simply would not be able to connect.
Fortunately Cerf and others saw this bottleneck coming. In 1996 they put together a new addressing protocol, IPv6, with 128 bits of address space. That means IPv6 can accommodate 340 trillion trillion trillion addresses. That should be enough for a while.
But the transition to IPv6 has been slow, as many organizations hesitate to make the needed efforts. Comcast IPv6 architect John Jason Brzozowski estimated that Comcast is seeing about five percent of users able to support IPv6 right now, though that number is steadily rising.
Today’s public moves by major websites like Google, Yahoo and Facebook, along with ISPs like Comcast and Time Warner Cable to completely switch to always-on IPv6 operations represents the first big addition of IPv6 connectivity since the protocol was launched in 1996. Joined by networking vendors such as Akamai and Cisco, this year’s efforts will begin to implement IPv6 broadly while keeping IPv4 connectivity on in parallel. Internet users, regardless of their connectivity status, should not notice any changes to the way they venture around the Internet.
Given that so few are even trying IPv6, why is it so important to adopt it?
In addition to making it possible for more devices to connect directly to the Net, faster and more granular connectivity could be another major benefit. IPv6 is not inherently faster, but because of the increasing shortage of device addresses, said Cisco fellow Mark Townsley, right now many devices have to aim their connectivity to other devices through the cloud.
As IPv6 becomes more widely adopted, Townsley explained, individual devices will be more able to directly connect to each other, without having to depend on the cloud as intermediary.
Forget 5.5 billion mobile devices. Imagine the possibility of billions, even trillions of pieces of hardware connected to the Internet, all sending out signals as simple as “this pen is out of ink” or as important as “someone is having a stroke that they don’t feel yet.”
Vendors are already lining up to create devices that leverage this capability. For example, More Than 50% of Devices at CES Were Internet Connected. Some hardware vendors, like Cosm (formerly Pachube) live in that space right now – helping device makers create devices that communicate efficiently and reliably on an increasingly crowded Internet. Because of its very mission, of course, Cosm has been IPv6-ready for quite some time.
Traffic Jams Ahead?
The picture is not all roses and sunshine: with the increase of devices will come an accompanying rise in traffic. Looking at the kind of data generated by the Internet of Things, many devices send small packets of data which shouldn’t overwhelm data networks. But as smarter devices handle larger pieces of unstructured data – and video streams – network saturation could become a very real problem.
Fortunately, we won’t have to deal with it all at once. In the Google Hangout, Cerf emphasized that the introduction of IPv6 is not so much a switch, but a transparent adoption of IPv6 connectivity as time goes on. And Google IPv6 engineers Lorenzo Colitti and Erik Kline added that since many networking devices are starting to be sold with built-in IPv6 and IPv4 features, the change to IPv6 is often technically not very challenging.
At Google, Colitti explained, “the problem is very wide, but not very deep.” Meaning that while he and Kline would have to find any software or device within Google that depended on IP addressing, actually flipping the switch once the code or device was identified was relatively simple. Both engineers expect that other organizations will have similar experiences.
The good news is that IPv6 is in place now and will be ready for organizations to use as they move to it. There’s no Y2K deadline of doom hanging over our heads. But companies looking to establish deeper connectivity with customers – and especially those planning to connect large numbers of mobile devices and IP hardware and applications - should consider beginning a gradual transition to the new addressing protocol.
Or you could simply wait until you find your percentage of IPv6 traffic. The nice thing to remember is this: virtually all networking software and devices will be able to handle both protocols for some time to come.
Source: Cisco All But Kills Cius Tablet
With the lackluster first day issue of Facebook on Friday, we thought we would take a moment to look at the memorable tech IPOs of the past and see how they have fared over the years. While the first day “pop” of some companies can generate news, what is more important is the longer-term performance of the stock – say, after three years of trading. The chart above shows some of these percentage gains – and losses.
One of the more memorable first-day increases was the doubling of share price for Netscape Corp. when it went public back in 1995 to raise the then-unheard-of sum of $1.6 billion. That’s less than 2% of what Facebook raised on Friday. Akamai raised twice what Netscape did in 1999 and had an increase of more than 450% in its first-day share price, only to fall to Earth three years later (thanks to the tech bust) and trade at 1% of its offering price. Ouch. Even Apple had “only” a 31% increase when it raised $3.4 billion in 1980 at its public offering. Three years after its IPO, it was down 25%. (Now is another story, of course.) And Paypal, which made billions for its owners and spawned an entire ecosystem of startups, was trading flat from its IPO three years later.
The biggest three-year percentage increases of popular tech stocks were Yahoo, with a 3,500% increase from its IPO, and Amazon, which rose more than 2,700%. Both benefitted from being on the right side of the tech bubble when they went public. Yet Yahoo has had its problems, as we have documented in the past. The last time its stock price broke 100 was at the end of 1999, and it hasn’t been anywhere in that neighborhood since then (right now, it is trading in the teens). Certainly, tech stocks are hot right now, and many – apart from Cisco and Microsoft – are close to their all-time highs including IBM and Amazon, both trading around 200.
One interesting trend not shown in the numbers is that all of the recent tech IPOs have come into the public markets with dual classes of stock shares, meaning that the public shares carry less voting rights (or in some cases, absolutely none) when it comes time for that annual shareholder meeting. Google will have three classes of shares next month: one for the founders, one for the public and then one with absolutely no voting power whatsoever that it will issue for dividends, employee incentive plans and acquisitions. LinkedIn, Zynga, Groupon and Yelp all have dual-class shares. James Surowiecki, writing in the New Yorker, says that this may “make the stock market less central to American capitalism.” A sobering thought indeed.
So what can we learn from this trip down memory lane? Just this: The first three years of a public tech company can be very chaotic times with regard to its stock. Some of the most successful companies didn’t make much money for normal shareholders. So if you are going to invest in a startup, start early, in its pre-money stage. Of course, that is pretty risky, too.