© 1998-2017, Arnold Consulting Group, L.L.C.
Offices in Washington, DC
and Vienna, VA


July 22, 2011

Why Apple might want to do the Hulu

Filed under: Apple,IBM,Technology — Peter Arnold

Why would Steve Jobs be interested in Hulu? The answer may be in Tarheel Country, specifically Maiden, near I-77.

That’s where you’ll find Apple’s mammoth new data center, built to handle iCloud and a lot more.  How much more?  Look at some numbers: This center is either 500,000 square feet (AppleInsider) or a million (Robert Cringely). By comparison, according to Cringely, IBM’s Special Events Web Service, which handled data for the Olympics, has three data centers with a combined 2000 square feet.

Any way you calculate it, iCloud by itself will not generate the data needed to give Apple anything approaching a suitable return on capital.  But wait!  Isn’t Apple trying to fast track the consumer transition to the cloud by phasing out (or minimizing) internal hard drives?  The new Mac Mini, introduced this week, has no optical drive and neither does the Mac Air, which replaces Apple’s aging MacBook.  Meanwhile, you’ve never been able to play a blu-ray disc on your Mac.

But that’s still not nearly enough data to pay for the Maiden center, especially given Moore’s Law (“the number of transistors on a chip will double about every two years”).

Let’s go one step further and factor in compression.  Even aging microchips in the Apple line have an H.264 encoder/decoder that apparently can compress a 1080p audio/video stream into four megabits per second.  That compares with about 20 megabits in normal circumstances or perhaps 24 for better HDTV.

Here’s the short answer on Hulu: Apple has pumped a huge amount of money into North Carolina but its returns, even if iCloud takes off, will be paltry-to-negative.  Even if you forget the money (Apple is sitting on $76 billion, after all), the black eye for the company would be a huge embarrassment.

So Steve and his colleagues need to do something fast.  Buy Hulu and integrate it into iTunes.  It’s not a great solution but as Richard Dreyfus demanded from Roy Scheider in Jaws, “You got any better ideas, hot shot?”

June 22, 2011

Once more unto the… cloud?

Filed under: Dartmouth College,IBM,Technology — Peter Arnold

Apologies for the long drought in blog updates.  Aaron Burr once said, “Delay may give clearer light as to what is best to be done” which may explain why Alexander Hamilton took the first shot in their duel.  But it’s not great advice for blogging.

Two events in the computing world this past week show that Serenditpity is more than just a well-known Manhattan eatery. IBM commemorated its 100th anniversary and Google issued its first foray into consumer hardware since last year’s Nexus One disaster.

First up, IBM.  Thirty years ago, Bill Gates offered to sell IBM the rights to the precursor to the Windows OS for about $100,000.  IBM refused, not comprehending the idea of profits in software outside its OS/1.  The driving principles in IT for the next generation offered a cruel lesson. By the late 1980s, minicomputers had pushed aside mainframes since VAXes were about a third the cost of an IBM 3090 and smaller too.  By the mid-1990s, PC’s had eclipsed minicomputers since they were about a third the cost of a VAX.

IBM’s saving grace was bringing in Dartmouth grad Lou Gerstner in 1993, whose approach to business (and himself) was so amoral that he had served on the board of a well-known cancer advocacy group only to resign to become head of RJR Nabisco.  But Gerstner understood computing’s changing dynamics, quickly reversing John Akers’ disastrous idea of quasi-autonomous divisions pursuing conflicting goals.  Instead, Gerstner killed OS/2 and turned the company’s focus to high-margin client services. (By 2004, his successor spun off the PC business to Lenovo.)

So here’s where tomorrow’s fun starts: As Walter Mossberg writes in today’s Wall Street Journal, Google’s new Chromebook computers are:

“essentially full-screen Web browsers designed to do everything via the Internet. Instead of using traditional programs, you will rely on ‘Web apps’ accessed through the browser—email programs, word processors or photo editors, for example.”

Among the benefits, notes Mossberg:

Because all your apps, settings and files are stored in the cloud, if you lose your Chromebook, or wish to use someone else’s Chromebook, you can just log into your Google account and all your stuff will appear on the new machine.

Google automatically updates the operating system, so you don’t have to deal with manual updates.

Look at the numbers driving this and the logical outcome.  Unlike certain industry observers, numbers don’t lie.  Depending on a company’s infrastructure (OK, a big “if”), cloud-based computing costs are at least a third cheaper than the cost of a PC, once you include the time and cost of customizing software, keeping it current, and adding security.

If the mainframe-to-VAX-to-PC experience holds, within another five years, the savings will approach 80 percent.  So if you’re in college now with your heart set on joining an IT department, your odds are about as good as beating the house at Caesar’s.

By that time, server farms such as what Apple is about to bring online in North Carolina to power iCloud will have proliferated globally, as will fast, low-cost M2M wireless connections.  At that point, your data is going into the cloud, backed by 256-bit encryption which isn’t perfect but is still far preferable to today’s system.

The Chromebook will not be a commercial success because consumer habits, especially in technology, are not conditioned for changes this radical, unless the company peddling it is based in Cupertino.  But cracks are in the Titanic’s hull and water is seeping in.  In another five to six years, Microsoft’s Office and Windows will evoke memories of Ozymandias, which is too bad insofar as Windows 7 is actually a worthy competitor to Snow Leopard.  But it is too little, too late.