Silicon Insider: The Last Days of Software?

Oct. 5, 2006 — -- Almost everybody in the modern world has heard of Moore's Law. Indeed, as the underpinning of the digital age, it is the metronome of our daily lives.

But very few people have ever heard of Moore's other, "Second" Law. And in the coming years it may prove to be even more important than the original. In fact, if the two laws continue forward as they have, there could be disaster looming for software giants like Microsoft.

The pace of hardware innovation, the central theme of Moore's Law, continues to surge forward while innovation for the accompanying software lags. It's probably one reason Microsoft has had so much trouble bringing it's much-delayed new operating system to market -- a development that has hamstrung computer-makers like Dell, not to mention the businesses and consumers who use their products.

Looking ahead, the innovation disparity could negatively impact not just software and computer-makers, but also the technology behind products like airplanes, cell phones, televisions and even MP3 players -- the things most consumers depend on every day.

Already, it is estimated by the industry analyst Venture Development Corp. that half of all new device projects now run behind schedule, with an average overrun of four months. That's going to get worse -- which means that the pace of innovative new products we've grown accustomed to (and depend upon in the case of medical diagnostic and treatment tools) is going to start slowing down… at the cost of comfort, entertainment and even health.

The 40-Year-Old "Law"

Moore's Law -- for our purposes, we'll call it Moore's First Law -- is taught these days to school kids. As you'll remember, it states that the performance (price, speed, size) of semiconductor chips doubles every 18 (these days, about 24) months. It was devised by Gordon Moore, co-founder of both Fairchild and Intel, in 1965.

The prediction Moore made in '65 not only proved correct, but, with a few modifications, continues to this day -- and the latest industry predictions are that it will go on to define the semiconductor industry for another 20 years or more. The "Law" has proven true not just for memory chips, but microprocessors as well -- and in almost every larger system in which these devices are used, from cell phones to personal computers, to even the computational areas (such as bioinformatics) of genetics.

It can be truly said that Moore's Law has been the defining material force for most of our lives, whether we know it or not.

So What About the Second Law?

Little known is the fact that the same paper in which Gordon Moore described his first Law, there is a short paragraph that, though it lacks the clarity and profundity of what came before, still effectively describes a second law. It goes like this: " …it may prove more economical to build large systems out of smaller functions, which are separately packaged and interconnected. The availability of large functions, combined with functional design and construction, should allow the manufacturer of large systems to design and construct a considerable variety of equipment both rapidly and economically."

About a year ago, I reintroduced this paragraph to the world in an article in Wired magazine. What I said at the time was that the underlying message of this rather obscure paragraph dealt with the efficiency of electronic devices. And just as Moore's First Law wasn't really a true law, but a promise by chipmakers to their customers to keep progressing the technology at a certain rate as long as it was physically possible, so too did Moore's Second Law seem to be calling for a comparable promise:

Overall net efficiency of any electronic system will double every 24 months.

Though the Wired article dealt with the desperate need for better batteries for electronic devices (a world which now seems to be unconsciously subscribing to this promise), I've come to realize over the intervening months that Moore's Second Law has even more profound things to say about other parts of the high tech world.

For example, the last year has seen a massive tectonic shift in the semiconductor business from single, monolithic processors to so-called 'multi-core' processors -- a shift that has left once-dominant Intel in trouble vis-à-vis upstarts AMD and Samsung. Why did these companies largely abandon the architecture that defined the industry for 40 years and was the heart of billions of chips? Because Moore's First Law snubbed up against Moore's Second Law: big, single processor chips had become too expensive, too inefficient. They were better capable of keeping up with Moore's Law, but had become too cost-inefficient to keep pursuing.

But this upheaval in the semiconductor business pales next to what may be a disaster looming in the world of software.

Did I just say 'disaster'? Well, look for yourself at the numbers.

Software Must More Than Keep Up with Hardware

A rule of thumb is that every time chip performance doubles, the software needed to tap into this increased performance quadruples. And if, according to Moore's First Law, chip performance makes that leap about every two years, then the amount of code has to double almost annually.

Unfortunately, that's just not happening. Industry analyst VDC estimates that the amount of code being produced in one of the best tech industries, telecommunications, is growing at only half that rate -- that shortfall helping to explain why the average software project in telecom is now six months behind, and getting worse. Look around and you can see the same problem almost everywhere. Today's average mobile phone has 2 million lines of code -- by 2010 it will have 10 million. The amount of code in the average new TV doubles, like Moore's Law, every 24 months. And this year's model cars will average 35 million lines of code -- a number that will jump to 100 million by 2010.

The great untold tech story of our time may be this growing shortfall between the hardware created by Moore's Law and the software needed to run it. It is certainly playing a part in the most quixotic business story of the last year: why giant Microsoft, whose very existence depends upon its ability to design and sell operating system software, can't seem to get its next generation OS, code-named Vista, out the door. The delays in delivering Vista have been like an anchor around the neck of the big PC companies, especially Dell. Yet, Microsoft keeps announcing one delay after another.

But the biggest software disaster of all may be waiting around the corner in critical industries we hardly think about as code-driven. It is estimated that 90 percent of all software code written is for so-called 'non-PC' applications -- from MP3 players to toasters to hospital equipment to passenger airplanes. If software starts falling behind in these businesses, then the pace of innovation itself -- something we've grown accustomed to over the last half-century, and upon which most of our business and economic models depend -- will begin to slow. More and more new products will be delayed, and those that do make it to market will be less reliable and trustworthy.

We aren't there yet. But that growing gap between hardware and software -- between our nearly religious adherence to the pact implied by Moore's First Law and our unwillingness to take on the equivalent commitment implied in the Second Law -- is becoming increasingly worrisome.

And it is going to get worse -- fast.

No Tme to Waste

Even as the demand for new code is growing by nearly 50 percent per year in some markets, the number of new software code writers out there is only growing by 8 percent per year. And it's not simply a matter of adding an army of new programmers in Bangalore either. Most of those folks, while good at testing and maintaining code, aren't prepared to develop high-powered, original new applications. That kind of work is still done mostly in places like Silicon Valley -- and there just aren't enough of those guys to go around.A few industry figures have seen this future and have become obsessed with finding solutions. Johan Wall, CEO of the Swedish/Silicon Valley company Enea, is trying to drive the device market towards a commercial off-the-shelf (COTS) standard. Others include Pekka Lettijeff, senior VP at Nokia, and Edward Screven, chief corporate architect at Oracle.

They all seem to agree that what it will take is a combination of more high-end code writers, more industry standards (so companies aren't working at cross purposes), more sophisticated techniques for recycling and reusing existing code, and perhaps even some major government-backed industry initiatives.

We don't have any time to waste. Moore's First Law is continuing to click along, powerfully and relentlessly, offering the hope of even more technological miracles. Now it is time for the tech world, and software in particular, to sign on to the compact of the Second Law as well.

Tad's Tab: The latest from the teen tech trenches, Malone's 15-year-old son, Tad Malone.

I like to joke that Facebook.com is for the people with goals in life, while MySpace is for those who don't. A more accurate division is that the former is used by college students, and the latter by high schoolers -- or at least those who share those values. After I quit MySpace, my friends turned me on to Facebook, which I have found to be more personal and better networked -- and with fewer weirdos and trolls. Nowadays, I can easily talk and share photos with my friends across the globe. For me, FaceBook is the future of social networking. It is living up to its hype

This work is the opinion of the columnist and in no way reflects the opinion of ABC News.

Michael S. Malone, once called the Boswell of Silicon Valley, is one of the nation's best-known technology writers. He has covered Silicon Valley and high-tech for more than 25 years, beginning with the San Jose Mercury News, as the nation's first daily high-tech reporter. His articles and editorials have appeared in such publications as The Wall Street Journal, the Economist and Fortune, and for two years he was a columnist for The New York Times. He was editor of Forbes ASAP, the world's largest-circulation business-tech magazine, at the height of the dot-com boom. Malone is best-known as the author or co-author of a dozen books, notably the best-selling "Virtual Corporation." Malone has also hosted three public television interview series, and most recently co-produced the celebrated PBS miniseries on social entrepreneurs, "The New Heroes." He has been the ABCNEWS.com "Silicon Insider" columnist since 2000.