September 16, 1996
An analyst report published today predicts that Microsoft and Intel, two powerhouses created by the explosion of the personal computer industry, face a troubled future where their growth and profits - even their viability - will be under severe threat.
Comparing Microsoft's plight to that of an animated character in a computer game frenziedly running for survival across a bridge while the bridge collapses behind it, the report's author, Robin Bloor, believes that we are in the midst of a seismic shift in the global $650 billion computer industry.
"This earthquake will totally rearrange the IT landscape, affecting all vendors of IT products and services as well as their customers," he says, predicting that the impact will be clear by autumn 1997 when the PC market moves into steady, irreversible decline.
The report paints a picture of a corporate computing world which is evolving toward a centralised network computing model, many times more complex, powerful and flexible than the former mainframe paradigm.
In this picture, demand for PCs declines dramatically to be superseded by stripped down, inexpensive PCs - called "thin clients" - with highly functional but very cheap browser front-ends (replacing the complex Windows-like operating systems that prevail today). These will be linked to a few powerful, massively parallel processing computers - called servers - all running Web-enabled software applications. These applications will be sold via the Internet, not today's third party dealers.
The main impetus for change is the substantially reduced costs delivered by the new model compared to PC computing. This could reduce the per user cost of computing by up to a factor of five. Bloor argues that the epicentre of the earthquake is the emergence of Java., a programming language from Sun Microsystems, that was designed for thin clients.
"Java will turn the world wide web into a vast network for client/server operation" Bloor says. "it allows software to be moved from a server (where it is stored) to a client, where it is executed with a browser. Under this definition, everything is software; a video is software, a sound recording is software, an electronic publication is software and an application is software. Because Java executes anywhere, for the first time software application development can take place separate from an operating system. This enables the Internet, changes the client/server paradigm and smashes the door open for the thin client."
Winners and losers
The impact on IT vendors will be felt most keenly by those with the greatest stake in the PC status quo, namely Microsoft and Intel. By contrast, the winners look set to be vendors that have traditionally operated at the higher end of the market, companies like IBM, Sun Microsystems and Hewlett-Packard.
Bloor identifies Microsoft's plight as a combination of several factors:
- the erosion, due to the emergence of the thin client mode, of its de Facto monopoly of the PC software market for applications and operating systems;
- slow adaptation to the emergence of the Internet and Java in particular which has put development tool revenues under threat;
- doubts about the viability of its strategy at the server level where reservations about the scaleability of Windows NT and the company's sales channels are growing. The report asks the direct question: 'Will NT scale (as it must to succeed in the emerging massively parallel computing environment)?' Bloor has technical reservations about NT's scaleability and has announced a forthcoming study into the scaleability of operating systems, including NT and believes the market for NT may falter as the industry centralises.
The report concludes: "Unfortunately for Microsoft, its desktop revenues, its development software revenues and its server revenues have all come under threat simultaneously. Thus it has to run fast to address this - it not only needs to preserve them, it reeds to keep them growing. We do not believe that Microsoft is capable of achieving this."
The report sees Intel as prospectively even more damaged by the emergence of this new form of computing than Microsoft. Sales (and prices) of its x86 chips will plummet because "the thin client does not need Intel thereby levelling the playing field". With Java dominant, this eradicates the lock in from which Intel has made such high profits. The chips used for thin clients will be cheaper and easier to program than Intel's current microprocessors.
The report argues that IBM is well-positioned to take advantage of the move back to a centralised form of computing. The company is still the prime vendor for many large corporate customers and centralised solutions already form the bulk of IBM's revenue stream. The company will probably take advantage of enthusiasm for thin client hardware and massively parallel servers. It is also likely to experience a boost in sales of corporate software, particularly of middleware, the integration software essential to making big networks, centralised or not, work effectively.
Bloor also believes Sun Microsystems will profit from the new World, not least because as the inventor and developer of Java, the company has won the hearts and minds of software developers around the world. As Web-enabled applications based on Java proliferate, moreover, Sun's presence in the Internet/intranet server hardware business is expected to mushroom.
The report predicts positive times for Oracle and Netscape as well. The report declares that the thin client will take off rapidly in the corporate environment from the Spring of 1997. The impact on vendors will become apparent by the final quarter of that year, by which point the PC market will have moved into decline.