杏吧原创

IT IS less than fifty years since William Shockley, John Bardeen and Walter Brattain invented the transistor, and less than forty years since Jack Kilby and Robert Noyce integrated multiple transistors on a silicon chip. From these two inventions sprang an industry that is beyond question the most important of our times.

In 1994, the semiconductor industry鈥檚 sales soared past the $100 billion mark for the first time. Chips are the sine qua non of all computer and communications systems, as well as consumer products such as compact discs. On them depends what is now the world鈥檚 largest and fastest growing industry, electronics.

But the speed and scale of its commercial success is not the only reason why the chip industry should fascinate observers. Where else do you find fundamental physicists rubbing shoulders with applied engineers? In what other business do sales staff have doctorates? How strange, then, that despite the semiconductor industry鈥檚 enormous economic significance and intrinsic interest, so little has been written about it.

Only a handful of books have been published, few of which have been worth reading. Most of the good books 鈥� for example, Mike Malone鈥檚 The Big Score, or Tom Reid鈥檚 The Chip 鈥� are now out of print. About all that is currently available, and worth reading, is George Gilder鈥檚 highly recommended Microcosm: The Quantum Revolution in Economics and Technology, (Simon & Schuster, 1989). So, in principle, any addition to such an underpopulated shelf is welcome, especially one that boasts a Japanese co-author who happens to be 鈥� although the book does not mention this 鈥� general manager of the semiconductor division at Hitachi, one of Japan鈥檚 largest electronics companies. For if the semiconductor industry in general is underreported, then its Japanese component is almost a black hole. And this gap is odd since, per capita, Japan produces more than twice as many chips as any other nation.

Unfortunately, the opportunity to illuminate goes begging. Tsugio Makimoto鈥檚 contribution to the book appears minimal. This is a shame, because he undoubtedly knows much that could have made enlightening reading. Not least would have been the inside story of how two young engineers under his supervision made key improvements to the process technology that subsequently became the mainstay of the chip industry. This is mentioned in passing as 鈥渁 major contribution in the 1970s which made possible such 1980s products as portable computers and camcorders鈥�. And that, frustratingly, is that.

Beyond such tantalising references, the book contains relatively little information about Japan, none of it new, and much of it wrong. To take just one, egregious passage: 鈥淎fter substantially re-engineering the Bells Labs transistor, in the course of which a Sony engineer called Leo Esaki discovered a phenomenon called 鈥榚lectron tunneling鈥� that won him a Nobel Prize 20 years later, a transistor was produced that was suitable for a portable radio.鈥� It was a General Electric transistor; Reona 鈥� as he then was 鈥� Esaki did not 鈥渄iscover鈥� electron tunneling: his Nobel prize came 16 years later, and the diode that he invented proved unsuitable for transistor radios, which is why Esaki left Sony to join IBM.

Makimoto鈥檚 co-author, David Manners, appears to be British (no American would use an expression like 鈥渂ombed the price鈥�). Alas, Britain is not the best country from which to write a book about semiconductors. Things seem to have gone wrong with British efforts to establish a presence in the chip industry right from the start. For example, Manners rehearses (twice) the pathetic tale of how Geoffrey Dummer of the Royal Signals and Radar Establishment 鈥減roduced the world鈥檚 first model of a chip鈥� at an international symposium in 1957, a year before Kilby made his seminal invention. But the model was a 鈥渘on-working鈥� one, and why poor Dummer never managed to put his idea into practice we are not told.

This is not to say that individual Brits have not made significant contributions to the semiconductor industry. There are currently, for instance, at least two British chief executive officers of Silicon Valley companies (Wilf Corrigan of LSI Logic, and Rodney Smith of Actel). But Manners鈥� grasp of Silicon Valley and indeed, of the semiconductor industry as a whole, is unconvincing. To cite just one small example, he consistently abbreviates the name of the chipmaker Texas Instruments to 鈥淭exas鈥�, but the company is invariably known in the industry as 鈥淭I鈥�.

Living with the Chip divides intno three parts. The first, taking up roughly half the book, is a history of the industry. This is almost entirely a rehash of the work of previous authors (none of whom are credited: the book has no bibliography and lists no sources). It is also riddled with errors of the sort described above.

The second section is a roll call of semiconductor companies, with a brief introduction to each. The chapter begins by stating that the industry musters 鈥渇ewer than 200 [firms] worldwide鈥�, then, after listing some 75 firms, the authors conclude, with characteristic inconsistency, that 鈥渢here are at least a couple of hundred more around the world鈥�. This listing could conceivably have made a useful refernece, if the authors had taken the trouble to organise it in some conventionally recognised manner. But entries are arrranged neither alphabetically, nor by size, nor by geographical region.

The final part of the book purports to explain 鈥渉ow the chip affects your business, your family, your job and your future鈥�. What we actually get, however, is a mishmash of other people鈥檚 ideas, many of them silly ones. Sample: 鈥渨ith dozens of the new flat panel screens hanging on the wall, maybe all in nice gilt frames, you could have a constantly changing picture gallery 鈥� impressionists one day, Dutch Masters the next鈥�. This may well be the kind of thing that an unimaginative billionaire like Bill Gates can afford, but with 25-centimetre screens currently selling for several hundred pounds each, it is not likely to affect the rest of us, or our families, in the foreseeable future. The semiconductor industry deserves better than this. Publishers please note.