WHEN PRODUCTS RULED THE EARTH
President Truman paraphrased an old truism when he once stated. "The only thing new is the history you haven't learned." Within the history of technology development, a cycle of initial complexity, subsequent simplification, and eventual reappearance of complexity has often been seen. This pattern is evident in the history of the automotive industry. The semiconductor industry also is at a particular point in its cycle, with complexity set to make a comeback.
The thesis is that for any given technology revolution, its beginning is announced by the appearance of product-oriented companies serving narrow niche markets. Consolidation follows, mostly doe to market and financial pressures, which results in large conglomerates offering generic products. Eventually, though, these conglomerates do market extremely diverse product lines. They do so by use of sophisticated process, rather than product, technologies.
When the automotive industry first appeared, the diversity of companies and products was bewildering. Each "instant" entrepreneur made cars for a niche market. Then, Sloan and Co. purchased many of these "silos" of excellence and consolidated their efforts into what is known today as General Motors.
Note the name. "General" products - single offerings - that attempted to satisfy the entire market. Simple automobiles designed to establish a market, rather than respond to it. At the time. market manipulation was the fashion; focus groups and market response studies were not done. The "Big Three" told the buyer what was needed, not the other way around. AS Ford stated, you could have any color you wanted, as long as it was black.
Today, no two cars are identical, at least potentially, because modern automation and information technology allow wide-ranging product variations. The manufacturer is closer than ever to responding to market demand in real time. Technology lends a greater voice to the user.
This is true in other industries too. Times turns many business concerns offering custom products into consolidated companies selling mass-production models. The same consolidated companies eventually offer many custom products. The circle is complete: today's factories are market-driven, not product-specific.
Semiconductor manufacturing is poised to enter this last, market-driven phase. Today's semiconductor facilities are product-specific and cost a billion dollars each. Think about it. A billion dollars is spent equipping a facility to meet the forecasted future demand for a chip it will produce.
But Moore's Law states that chip density increases by a factor of 10 every five years. We are now at 0.2 micron technology with 10-million-junctions-per-chip. In the early 1970's a megabyte of memory cost a half-million dollars. Now the cost is about forty bucks. How long can this go on.
Single-product semiconductor factory efficiencies are decreasing as the market develops, thereby delivering less bang for a buck. Semiconductor manufacturers must take a page from the discrete-manufacturing handbook: make agile factories that resist obsolescence. To take the example of the auto industry again, the trend is away from transfer lines to more agile CNC-machining centers automated with robotics handling.
The end results of agile manufacturing is that we can become managers of the present and not the future. Managing the future always causes headaches. Agility and flexibility are the watchwords of any commodity-like technology business, not any one specific product. The semiconductor factory of the future will be centered around the process of manufacturing, not the specifics of the product.
The transition will offer many opportunities to process-equipment vendors. Rather than the academic-oriented equipment found in existing semiconductor facilities, Silicon Valley will engage in agile execution similar to that found in other maturind industries. This will reduce the cost of new products. Wall Street valuations placed on semiconductor vendors will drop to reflect realistic expectations of revenue and margin, not hype. Silicon Valley will learn from the Rustbelt.
As appeared in Manufacturing Systems Magazine February 1996 Page 84
References - Table of Contents
|R. Morley Incorporated
614 Nashua Street, Suite 56
Milford, NH 03055-4992 USA
Tel: 603-878-4365 FAX: 603-878-4385