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

<< [ Get a Life (Which One?) ]   |  Upgrade Denied  |   [ Resistance is Futile ] >>

Weekly Column

Upgrade Denied: If Comdex is a Drag and the Christmas Quarter is a Bustfor PC Companies, Maybe We Just Don't Need That New PC

Status: [CLOSED]
By Robert X. Cringely
bob@cringely.com

This is the week for Comdex, the Computer Dealers' Exposition in Las Vegas,usually referred to as the largest all-computer trade show in the world. Butthe economy is down and fewer companies are exhibiting, so maybe this yearComdex won't be the biggest, but who cares? Twenty years ago, the big show wasthe National Computing Conference, and it no longer even exists. Trade showsjust aren't something to get emotionally attached to — it's JUST a darnedtrade show. Now tractor pulls, THOSE you can get attached to.

A good question to ask right now would be, “Why Comdex is declining?” The reasons are more complex than just the bad economy, because bad economies have not always affected the Information Technology business like the rest of theworld.And the reasons you think IT might be on the ropes might not be thereal reasons at all.

Every IT sector is down dramatically from highs generally set in 2000, andwhile it is easy to talk about the dot-com meltdown and CFO shenanigans, thosecataclysmic events aren't the primary causes for this continued decline. Thereal reason IT is in trouble comes down to a product upgrade cycle, or rather,the lack of one.

Moore's Law, defined by Intel co-founder Gordon Moore as the doubling ofcomputing power at no increase in cost every 18 months, makes every newgeneration of computers faster than the one before. And since computer nerds,who often govern when new computers are needed in an organization, aregenerally unwilling to be more than a single generation behind, that has longmeant that everyone gets a new PC, your department gets a new server, and thecompany scores a new mainframe — or a significant hardware upgrade to thecurrent mainframe — every three years. This is the upgrade cycle: Everyonegets a new machine every three years. The personal computer industry inparticular is dependent on this cycle, because there are almost no businesscomputer sales anymore that aren't replacements. Every desk that can holda PC already has one. Every department already has a server. Mainframe computer rooms are getting smaller, not larger, and outsourcing is limiting newsales of big iron machines.

So everyone usually gets a new computer, but not this year, and probably notnext year, either. For the first time ever, it looks like U.S. industry mayskip an entire upgrade cycle, which has a harrowing effect on companies likeIBM, Hewlett-Packard and Apple. Part of the reason for this skipped cycle isthe bad economy, but bad economies never caused an upgrade cycle to be skippedbefore. Another part of the reason for the skipped cycle is Y2K.

On the way to spending more than $100 billion to make sure our factories and water systems and credit cards and nuclear power plants didn't stop working atmidnight on December 31, 1999, almost everyone in every business got a newcomputer. The year 1999 was a very good year for computer hardware andsoftware sales of all types as Y2K was used as the excuse for buying new stuffin just about every category, whether it was really needed or not. Who knew?And remember that Y2K was right at the end of the period in which industry wasalso adjusting to the Internet, so lots of more powerful new computers andnetwork gear were purchased so everyone could surf the Net, too. Y2K and websurfing worked together to put a stutter step in the next upgrade cycle thatmight have happened in 2002 or 2003, with the result that not many businesscomputers need replacing right now.

It could be argued, in fact, that no computers at all need replacing and thisis the third reason for the skipped cycle. This time — at least on thedesktop — blame Intel. Locked in a death struggle with the scrappy AdvancedMicroDevices, Intel several years ago decided to invest in parallel lines ofprocessor development so that Moore's Law could be artificially acceleratedeven faster. The idea was to outspend AMD just like Reagan outspent theRussians and thereby put AMD permanently out of the game. It hasn't quiteworked that way, but the result has been a cascade of new Pentium chips, eachfaster than the one before, and all of them faster than most users actuallyneed.

Intel's headstrong rush toward the future means that hardware is now generallyahead of software, and the time it takes your new PC to recalculate aspreadsheet isn't discernible from the time it took your old PC to do the samejob. This is not good for upgrade sales. Only gamers and people who runadvanced 2D and 3D graphics programs for jobs like TV animation and specialeffects even tax their computer hardware, which makes it difficult to justifyasking for that new PC.

As a result, the hardware upgrades for business that might have happened in2001 and 2002 have essentially not happened, and it doesn't look much betterfor 2003. Computer industry analysts generally look for flat to slightly higherunit sales in 2003 over 2002, which isn't good news for an industry where theaverage unit price is dropping at the same time.

The only good news in PC hardware has been the home market, which is nowapproximately 35 percent of total PC sales and expanded giddily in 2000 and2001 as average folks discovered the Internet and then discovered that theirold PC wasn't fast enough for the Internet. Alas, those surfers are allgenerally now onboard with new hardware, so home PC sales for 2003 don'tlook to be that good, either.

Software follows hardware as people buy new software to run on their souped-up new computers. No new computer means no new software, or at least not as much new software, so that business looks soft, too.

So where is the good news for these companies? There ARE some bright areas,mainly wireless networking and video games. But WiFi networks have alreadybeen commoditized, so while lots of companies will make money with them, fewcompanies will get rich. And video game machines, well, the manufacturers losemoney on those, so booming sales don't mean booming profits until a coupleyears later, when royalties from further game sales really start to kick-in.

The best I can come up with for now is that most of the companies that willDie have already died. A few companies, notably Dell and Cisco, are using thedownturn to kill enemies and grab market share, which bodes extremely well forthem and extremely ill for their competitors. HP, for example, has had "adagger stuck in its heart" by Dell, according to a friend of mine on WallStreet. I have no doubt that Dell will eventually be the biggest computercompany on Earth, which doesn't make them the best computer company or themost original by a long shot, but buyers will know their warranties will behonored.



Comments from the Tribe

Status: [CLOSED] read all comments (0)