Thursday, April 21, 2011

Apple's new Data Center

Being an avid reader of Apple related news, I’m just fascinated by the function of the newly built back-up center in North Carolina. Yes, most of the articles in the net are usually covering predictions of iPad3 or iPhone5, their features, release dates, etc. However, I would say that the data center, though it is less sexy than the shinny polished gadgets that we can get our hands on, it is much more interesting and would have far reaching consequence to the Apple ecosystem than most consumers would think.

Before Apple made iPhone, people had predicted that Apple would make a phone. Before, iPad was release, people already know Apple was making a tablet. This time, when Microsoft, Amazon, and Google are drumming for the possibility of Cloud Computing, you think Apple will sit around and do nothing? Certainly not, I think the data center would the key for it.

Apple’s strategy (non-financial wise) is to provide the best consumer experience in using personal technology to interact with information and consume entertainment. Desktops, laptops, phone, tablet, etc take care the physical part, the OS, itunes, app stores, etc are the software part. Apple still has to work on negotiation with content providers to get their contents on Apple’s platform to be delivered to consumers. Cables will be gone, wireless is the future, that’s where the cloud computing come into place. I’m no expert on this subject, but I read the article this morning and just love the analysis of what Apple’s data center would do. Just wanna share with you all:

What No One has Dared Think About Apple’s Data Center
Columns and Opinions by Adam Christianson and John Martellaro 2:29 PM, Apr. 20th, 2011
“When you are solving a difficult problem, re-frame the problem so that your solution helps you learn faster.” — Aza Raskin

There is a classic puzzle about a fellow in a rowboat who is rowing down a river and drops his hat overboard. The puzzle involves calculating how long it takes him to turn around and row back to the hat floating in the water. The problem is a lot easier to solve if one changes reference frame and works from the guy in the boat instead of from the shoreline view. It’s an example of how a change of reference frame is often useful for solving a problem. Another example comes from physics. Some problems in electromagnetism are a lot easier to solve in spherical coordinates. In other words, recasting the problem to make it easier to solve is a helpful technique. The quote above for this article extends that idea to not only re-framing, but also doing it in such a way that we learn faster.

It’s that recasting that I want to talk about with respect to Apple’s data centers.
A lot has been written about Apple’s new data center in North Carolina, but no one has been able to put together a coherent picture of what it all means. Most articles wax into lofty notions of a massive data center to deliver a better MobileMe experience, untethered iPads synced to the cloud instead of a PC — or a massive movie databank. And leave it at that.

Apple’s North Carolina Data Center
The problem arises when we start to formulate an idea about just exactly how the data center (and possibly more like the NC center worldwide) would be utilized. For example, if one takes all the assumptions about Apple’s DNA, it’s easier to re-frame the question about what Apple trying to achieve.


Let’s start with some easy analysis about Apple’s competition, Netflix. If Apple intends to out-Netflix the Netflix Corp., then how will Apple convince Hollywood executives to give them rights to a massive movie archive as Netflix CEO Red Hastings has done? (We know that Hollywood executives are alarmed at the prospect of Apple obtaining control over movies in the way Apple seized control of music.) We also know that Hollywood and the networks don’t want to upset the apple cart with the lucrative deals they have with cable and satellite carriers.


We suspect that buying Netflix, with its huge inventory of plastic discs and likely exclusive contracts with Hollywood, isn’t Apple’s style. (That is, Netflix’s contracts might not be inheritable if the company is bought.)


In fact, everywhere we turn, it seems like the problem is insurmountable and in desperate need of reformulation. The whole scenario suggests that we’re asking the wrong questions, making the wrong assumptions, and, worse, not analyzing the problem that Apple faces and imagining a next generation solution.


Seeking an Analogy
A good analogy comes from looking at the music industry and the development of the Apple iPod. The music industry in the 1990s was going strong and was all too happy to sell complete albums on a plastic CD. And we know how much they charged for that piece of plastic which forces us to buy all the songs in the album. It also forced customers into ridiculous collections, bookcases, boomboxes, and awkward CD players. Apple came along with a 140 gram iPod that could hold 1,000 songs and that changed the game completely. The labels’ obsession with money and the status quo blinded them to the next technological step, which was user choice, playlists, and massive storage in your pocket.


Asking a New Set of Questions
Let’s ask some new questions and see where it goes.

1. What if Apple’s analysis of the TV industry has led to the same kind of ground breaking, imaginative solution?
2. What temptations could Steve Jobs lay before Hollywood and the networks that would allow them to maintain the kind of control they want yet lure them into experimenting with Apple at a higher level?
3. How could a massive data center become, essentially, the next (super) iPod?
4. Apple recently made headlines by purchasing 12 petabytes of storage. One of our readers calculated that 12 petabytes would take up about 180 sq ft of floor space. Apple’s data center has more than 500,000 sq ft. We’re talking exabytes of possible expansion. More as storage continues to use less space over the lifetime of the data center. What service could use that kind of storage?

Some Assumptions
Let’s summarize our reasonable assumptions about Apple’s situation.
1. Apple doesn’t want to buy Netflix because it doesn’t want to inherit a mail order business of plastic discs. It may not be able to inherit the movie rights.
2. Apple doesn’t want to buy an ISP like Time Warner or Comcast and inherit the lousy politics, bureaucracy and customer service reputation of these companies.
3. Apple can’t reasonably replace the broadband infrastructure currently in place by telcos and cable companies. Apple would have to spends hundreds of billions, as AT&T has, to build a new, national wireless network. That’s not a good use of Apple’s money.
4. The Apple TV has taught Apple a lot about the home TV industry. It remained a hobby because Apple hadn’t solved the basic problem at hand.
5. Apple doesn’t want to own the newspaper, magazine, book, music and movie industry. Instead, they want a 30 percent cut, a piece of the action, on the delivery of all that content.
6. By the end of 2011, Apple will have 50 million TV sets, called iPads, in customer homes.
Some Guesses


The questions above are a completely different set of questions than we’ve have seen asked, and while we don’t have all the answers, we’ll try to make some educated guesses based on the questions and the assumptions.


First, we note that here are too many fragmented delivery mechanisms owned by different players and which use different technical delivery mechanisms. And each has its own unique library and UI for delivering content. Recall how, in the early 1990s, many of us had cassete players, CD players and VCRs. Now, we have DVDs/Blu-ray, Internet (Hulu, Roku, Apple TV, Boxee, Google) and cable/satellite. Vested interests lure us into buying physical media that’s constantly being made obsolete. Vested interests are keeping the industry from thinking about how to deliver all this content is a new, more convenient and coherent way.


Our chief complaint is that American households are constantly changing between these delivery mechanisms to get a better price for bundled Internet service plus content as they also wrestle with new hardware and pray for better customer service. It would be better to take content out of the equation and reduce the carriers to fast pipes by offering customers a superior content purchase and management experience.


In other words, the problem is not the content. Many movies and TV shows are terrific — once they get delivered. The problem is all the fuss we have to go through to get that content delivered (and stored, and backed up). What we need is a better channel. (For example, who uses Epix? Who has even heard of it?) What Apple could be building is the channel everyone wants to be on.


So, like the iPod, it seems that the key to all this is a simple, easy to use, dependable, customer friendly channel that delivers the content that Hollywood wants to deliver. To do that, Apple needs a modern, capable data center.


For example, here’s how content delivery happens for many people now:
[Comcast (etc), DIRECTV (etc)] -> DVR -> HDTV

Here’s how Apple is slowly introducing an evolutionary new channel. Each step is part of Apple’s plan, but it’s been one step at a time.
Mac (iTunes) -> Apple TV 1G -> HDTV
Apple Cloud -> Apple TV 2G -> HDTV
Apple Cloud -> iOS device (AirPlay) -> Apple TV 2G -> HDTV
Apple Cloud -> iOS device (AirPlay) -> (optionally) HDTV

The last step suggests the elimination of the Apple TV, and that’s why we’ve been hearing about Apple making AirPlay available to the TV makers. (Even so, Apple can keep selling the Apple TV 2G as part of a stopgap household solution.) This evolution could also explain why we’ve heard rumors about Apple making its own TV: to seed customers with an AirPlay capable TV to trigger interest in the technology, not to globally compete with, say, Panasonic and Sony.


The overall idea here is to make the user experience with its family of iOS devices so superior that customers will naturally want to opt for the Apple channel rather than the old, obsolete technologies the cable companies have been delivering.The overall idea here is to make the user experience with its family of iOS devices so superior that customers will naturally want to opt for the Apple channel rather than the old, obsolete technologies the cable companies have been delivering. After all, these companies have been modestly successful delivering bandwidth, but have been terrible at delivering mobile technology devices with modern UIs. With hundreds of millions of iOS devices in consumer hands in the near future, customers will naturally opt for these devices rather than the sorry, disjointed services, UI and equipment the cable companies have been able to muster.


Apple hopes the content delivery services of the bandwidth suppliers will dry up and blow away thanks to customer demand for Apple iOS devices and technologies. At the very least, Apple will surge to the top when it comes to customer choice for delivery, making it very difficult for Hollywood and the studios to avoid doing a staggeringly large business with Apple.


The associated step is to point out to Hollywood and the studios that links to this content are everywhere. Every movie, every song, every book has a URL. The embedding of those URLs is everywhere in the web. Apple has 200 million customers with credit cards on file who are always just two clicks away from buying content from their iPhones and iPads, etc. That’s the leverage Steve Jobs will have, along with the insame popularity of the iPad, to lure content developers onto the Apple channel.


Apple will have bypassed the content from the cable companies without building its own network. And Apple gets 30 percent of all the action.
________________________________________
John Martellaro is a senior scientist and author. A former U.S. Air Force officer, he has worked for NASA, White Sands Missile Range, Lockheed Martin Astronautics, the Oak Ridge National Laboratory and Apple Computer. During his five years at Apple, he worked as a Senior Marketing Manager for science and technology, Federal Account Executive, and High Performance Computing Manager. His interests include alpine skiing, SciFi, astronomy, and Perl. John lives in Denver, Colorado.
You can follow John on Twitter at twitter.com/jmartellaro.

No comments: