The Industry Cools Down
Ephraim Schwartz is playing darts again – putting the right bulls-eye on the Nicholas Carr proposition. No – not the alarmist “IT Does Not Matter” notion but rather that the IT industry is cooling down. Or more pointedly consolidating. This differs from Carrs message in two respects. Carr was saying in effect in all the major markets that computing was becoming a way-down the learning curve commodity-pricing game . Secondly Carr posited a slowing in the sweep if not the rate of innovation and therefore the relevance of IT to business.
Carrs analysis fell far short in two respects. First, the position that IT systems, processes and computing devices have garnered over the past 65 years of waves of change is pervasive in business. Just the legacy technology alone has concurrent and late adopter changes to account for. Everybody in business knows that only the top 20-40% of companies worldwide are near the effective ROI edge of IT technology.Just look at BI and Content Management. This leaves room for huge “after” markets – and that without the kick-in of replacement after markets.
But even more telling, Carr failed to see the re-investment waves, just as huge as the basic Mooores Law driver of IT computing change. Embedded computing devices are now completely making over key IT infrastructure such as storage, communications, and now basic transportation (think RFID), making each of these pivotal industries subject to massive and explosive innovation opportunities. These in turn feedback to the business core and are causing major repercussions on the base infrastructure. Thus,post-Carr we are witnessing the sea change that is the center of IT universe moving away from the desktop and back to central servers. This is a yin-yank cycle over the 65 years of computing which sees servers become once again the most cost effective ways of controling and distributing the desktop experience and GUIs to literally hundreds of different smart devices. This is an unforseen, un-Carred … uncalled for sweeping change reaching its critical phase right now as RIA, SOA and Web Sevices take off.
Consolidation as Point of Convergence
Where Ephraim and Carr overlap is on the notion of consolidation in the industry. Major segments are standardizing. Take computing chips-bigtime cooldown if by “cooldown” is meant not number of devices used or even revenues; but rather the number of competing computing chip architectures. In mobile devices its Intel+ARM and everywhere else its Intel x86 with possibly one or two marginal in numbers, special server architectures like PowerPC or Intels own Itanium. Trouble is for Intel that the best innovations in x86 technology are coming not from Intel but AMD.
In software, Bill Gates would dearly like to see the same consolidation take place particularly in operating systems and desktop applications. And going on ten years after Windows displaced its most serious desktop competition – the much more reliable, scalable and secure OS/2, Windows Vista is finally zeroing in on delivering just such robustness. But WinPF/Avalon, WinCF/Indigo, WinFS/Yukon and .NET 2 still need to round into form over the next 1-2 years to finally deliver on the promises that were Windows 3.1 and 95 of a decade ago.
Meanwhile the endless parade of scalability then availability then bloatware and now security problems with Windows over the past ten years all done to a backdrop of constant reliability and interoperability shortcomings have left the market and IT consumers open to alternatives in OS and Office applications. On the server side, both Linux and Solaris seem to be seizing the opportunity and developping very loyal constituencies. On the desktop, Linux and particularly Mac are setting up for a conflict as Microsoft makes the critical transition to an essentially new and more worthy operating system. Linux brings lowest unit cost plus a much wider set of applications and 80-90% equivalent functinality while Mac brings a level cost-of- hardware playing field and
But in the database arena no contender among IBM, Oracle, SQL Server and OpenSource have been able to secure a dominant position; but just next door in ERP and business applications, Oracle may have gobbled up a enough competitors to put it in position to take on SAP and Microsoft for leadership in the core business resource management apps. Likewise, there is movement towards convergence in the Application server, BI and Content management markets but hardly as pronounced as the oligopolies emerging in some of the large scale markets. And nothing to match the pre-eminence of Intel in the computer chip markets.
So yes Ephraim is right. There is a cool-down and consolidation at computings infrastructure core – but it is on the chip architecture side, not on the software side. And all bets are off if somebody figures out how to put gesture or speech recognition on a chip. There are some major confrontations upcoming in the software arena … and in the rapidly evolving world of embedded and connected devices there is a whole new Schumpeter world of creative destruction just emerging – and there the binding OS may be a programming API.