Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
I, Cringely - The Survival of the Nerdiest with Robert X. Cringely
Search I,Cringely:

The Pulpit
The Pulpit

<< [ The Puppet Master ]   |  The Power of Six  |   [ I Report, You Decide ] >>

Weekly Column

The Power of Six: Google's plan for world domination. Also why the iPod Classic sucks.

Status: [CLOSED] comments (93)
By Robert X. Cringely
bob@cringely.com

I wrote a few weeks ago about Google’s attempt to influence the rules for redeployment of the 700-MHz radio band in the U.S. for voice and data applications. Google said it would agree to pony up the $4.6 billion auction reserve price if only the FCC would first guarantee to force any eventual winner to keep the frequencies open in a variety of Google-defined ways — ways that were decidedly unpopular with incumbent U.S. mobile operators. It seemed to me to be lunacy for Google to deliberately po the mobile carriers if it wasn’t going to spend the big bucks to actually WIN the auction. But what if Google DOES plan to spend the big bucks and win the 700-MHz auction? What would they do with it? I now think I know.

Google didn’t get what it asked for from the FCC, which opted for a different definition of “open,” promoted by the FCC commissioner. The commissioner’s definition of “open” was also opposed by the incumbent carriers and, in fact, Verizon is apparently taking the issue to court, but I think they doth protest too much. The carriers can probably live with the existing auction rules. Fighting them in court is intended as much to signal Verizon’s determination to win the auction as it is to actually overturn the auction rules. The last thing Verizon wants is for Google to enter the auction AT ALL, because doing so can have only two consequences, neither of them good from the perspective of the telcos: 1) Google might actually win the auction and impose the very rules it tried earlier to get with the FCC, and; 2) the mobile carriers might still win the auction but Google’s involvement would cause them to bid much more for the spectrum than they otherwise might.

Google could make it VERY expensive to hold together the existing U.S. mobile phone oligarchy.

Remember that none of the existing U.S. mobile phone companies is currently lacking in bandwidth. They would love to own the 700-MHz band if they can do so cheaply, but they don’t apparently have any real intention to USE it, which would mean building out a whole new infrastructure at the cost of several billion dollars. They just want to bank the spectrum and keep it away from Google.

It seemed to me that the greatest impediment to Google actually spending the big bucks to win the auction (they could clearly afford it) is that the mobile phone and data businesses aren’t as profitable as Google’s own search and advertising businesses, which means making such a move would hurt Google’s earnings and be a drag on the price of its shares. This seemed to be the difference between Google posturing and Google actually doing something.

But then this week Apple began to bluster about entering the 700-MHz auction, which makes even less sense. Could this have something to do with Google? Like a lot of other pundits, I keep facing the fact that Google CEO Eric Schmidt is on the Apple board and expecting that association to manifest itself eventually in some form of product or service alliance, but that has yet to happen. Could this finally be the time? Apple AND Google have together more money than anyone except God, Warren Buffett and Bill Gates. Could the two companies be intending a joint bid of such grand proportions as to guarantee a win? And if they did, what way could they find to use the spectrum that wouldn’t be a drag on Google’s earnings after all?

So I thought and I thought and I came up with what you are about to read. As usual this is just guessing on my part, but I’m a pretty good guesser.

To start, I don’t think Apple will actually bid with Google or even against Google in the 700-MHz auction. It would overcomplicate the five-year iPhone deal between Apple and AT&T - a deal that is already strained by the rise in third-party iPhone unlocking tools. (What was the chance Apple didn’t see those coming? Zero.) While it is possible that Apple would deliberately go against AT&T because Apple is, well, Apple and likes to stir things up, I think there are limits to how much Hell Steve Jobs is willing to raise in the wireless space given the string of global iPhone deals he is still putting together. And Steve is cheap, too, meaning that he might not see this as a good use for Apple’s free cash.

Besides, Apple has enough trouble on its hands with the new iPod classic, which doesn’t work very well at all and is going to shortly create some PR problems for Apple. Rather than actually being a legacy device as the name implies, the iPod classic uses new innards and the software is creating headaches for early users.

The complaints I am hearing about the new iPods classics are (in no particular order):

  • VERY Slow menu switching response
  • Display of clock rather than song info when “Now Playing”
  • Inability to use existing AUTHORIZED 3rd party dock products (including Apple-advertised)
  • Audio skipping during operation
  • Slow connection to Macs and PCs
  • Inability to disable “split-screen” menus
  • Lagging and unresponsive Click Wheel
  • Camera connector not working
  • Inability to use EQ settings without skipping and distortion

This product was clearly shipped before it was ready, so we can expect a significant firmware upgrade Real Soon Now, especially since the iPod classic is now Apple’s ONLY solution for users who want to store more than 16 gigabytes worth of songs, pictures, TV shows, and movies.

So Apple will have its corporate hands full between now and Christmas, which is yet another reason why I seriously doubt the company will be involved in 700-MHz auction action. Apple’s current rumblings about the 700-MHz band are more likely Jobs helping Schmidt. If the mobile carriers interested in the 700-MHz band think that it might cost them $16 billion rather than $6 billion to win the auction, they might not bid at all, allowing Google to get the property for less than it might have had to pay in a contested auction. At some price the deal becomes uneconomic for the mobile carriers and, given their small minds and squinty eyes, they’ll see it as uneconomic for Google, too. “Let Google take the fall,” they’ll think.

But Google won’t be falling.

The huge expense of buying the 700-MHz band and building out the infrastructure could be made a lot less huge if Google didn’t have to build out the infrastructure. No traditional mobile company could get away with this, but I think Google could.

I have written about nearly all the individual parts of this before and even wrote a column putting it all together, though as my idea, not Google’s. Maybe they have been reading me after all.

First let’s start by looking at the infrastructure Google has already built or committed to building — the largest fiber backbone in the world and the largest and most widely distributed data center build-out in the world. Both are FAR in excess of Google’s current or even future requirements UNLESS they are also intended to work with a massive 700-MHz wireless network.

Imagine a hybrid wireless broadband mesh network using 700-MHz connections for backhaul and some truly mobile links and WiFi for local service. Google has enough experience with WiFi in Mountain View to know that it isn’t, by itself, a good solution for wide area networks. The key failing of metro WiFi networks is backhaul to the Internet backbone. But if Google used its 700 MHz band for that AND implemented it as a true mesh network, there would easily be enough capacity to serve almost any size network given a suitable number of backbone connections.

You can find my old column about just such a network in this week’s links.

Google has experience, too, with hybrid wireless networks. Every Google employee has the chance to take a company bus to work and every Google bus has an EVDO-to-WiFi bridge so Googlers can surf the net on their way to work.

It would be really cool if this Google hybrid network was truly flat and could be maintained entirely within a single address space like, for example, the 76 billion billion billion IPv6 addresses Google already owns. The sudden existence of a massive IPv6 network would throw other ISPs into a tizzy and quickly drag the rest of the net into the 21st century, something else I could see as a Google ambition.

Finally, what links all of this together is something else I wrote about long ago — the Google Cube. This is an access device that contains 700-MHz and WiFi radios, a tiny Linux or Linux-likeserver, and a few gigs of flash RAM memory cache. It’s these Google Cubes that will mesh together, acting as both WiFi access points and 700 MHz mesh backhaul devices. Throw in some local caching, video preloading, and truly local DNS service and suddenly you have a pretty substantial network infrastructure that is not only massive and self-healing, IT IS ENTIRELY PAID FOR BY CUSTOMERS. All Google needs to provide are several thousand points-of-presence (cell towers) to connect the local mesh to the Internet backbone.

Google couldn’t do this with WiFi alone, but with 700-MHz meshing and backhaul they could make it work fairly easily and the entire network could be deployed in a couple months.

For those who can’t think past search, imagine this also as Google’s key to dominating local- and location-based search.

Forget about net neutrality and forget about making nice-nice with broadband ISPs OR phone companies. Google would overnight become the largest U.S. ISP with direct and very high-performance access to its customers, including those using the new Google Phone or any other phone that supports WiFi connections, like the iPhone and many others. Google becomes the biggest and lowest-cost ISP and potentially the biggest and lowest-cost mobile phone company in the bargain.

Heck of a deal.

Comments from the Tribe

Status: [CLOSED] read all comments (93)

Would you add into the mix Google's interest in FON, maybe as a kind of experiment?

Venkman | Sep 21, 2007 | 10:25AM

A shake-up in the industry has been needed for some time now.

Neil Anderson | Sep 23, 2007 | 9:04AM


Bob,

Please call up any wireless expert and ask them HOW MUCH data you can actually push over 700Mhz. You'll be surprised, it isn't a whole lot.

Irving | Sep 25, 2007 | 3:19PM