Nilesh Jasani's Reviews > Chip War: The Fight for the World's Most Critical Technology

I write this review a few days after the book's publication and the day after the US announced a sweeping set of restrictions on US tools sold to any advanced Chinese semiconductor chip manufacturer. There could not have been a more forceful endorsement of the relevance of any book.

Chip War is more than topical. It presents a highly readable history of the industry. While refreshingly unbiased in many arguments, the book sheds light on the thinking pervasive amongst US policymakers. In a world where most software (from the search to finance to consumer use related) is relatively easily replicable, most policymakers feel that the non-duplicable part of the technology industry is what one can do on a silicon wafer. The US believes that it is the leader in this segment - along with a few others who are a part of its influence sphere - and should use not just innovations but also other political tools to retain supremacy.

The author is clear in drawing his lessons from a chronological set of main events and personalities that built the industry in the West. The following are some of the questions not asked in the book but worth thinking about based on the same historical facts.

a. If one looks at the market caps of upstream chip manufacturers as a percentage of the total technology industry capitalization over time, semi-companies do not seem to hold as much monopoly power or importance in the eyes of investors as believed in many other circles

b. While the book mentions a good number of early innovators by name from the earliest decades, there has been almost no named scientist in the last twenty-five years (this is eons in the tech industry). There are several potential reasons behind here: narratives and tales surrounding innovators are far more common in the United States than in Eastern nations. Also, innovation stories are more relatable when they are easier to understand - most semiconductor manufacturing innovations for the last thirty years are in detail and not in broad subjects like transistor or UV lithography. And lastly, the latest tech innovations are team- and resource-based rather than individual-driven.

c. Semiconductor manufacturing knowledge is highly diffused. Many in the US may feel that the knowledge is all American because of the work of the earliest giants, but this is no different from if Europe were to seek credit for early innovations in internal combustion engines or nuclear theories. Even if some of the knowledge had been previously stolen, the most critical parts of advanced manufacturing now are not all necessarily in the US. The rise of the likes of TSMC and Samsung Electronics, along with the inability of American companies to keep pace, are proof. The diffusion of the know-how happened quite early as reflected in the dominance of the Japanese - well described in the book - in the 1980s.

d. Nobody has been able to maintain dominance in this space for long. The book is littered with the tale of well-known names that led the space for a few short years. From Fairchild, Texas Instruments, through Motorola, Nikon, and now Intel, one can name a couple of dozen companies that seemingly had an unbeatable innovation lead only to see someone else take over. Is semiconductor manufacturing so challenging to master that a lot of money and an army of engineers cannot solve it without being a part of an existing establishment?

e. One wonders if the military establishment needs instruments created by the most advanced fabs of this year? It is clear that most rival superpowers of the world have access to technologies of a few years ago - say 2014/5 - even if they are blocked on the latest fabs. If, in the worst case, either the US or China does not have access to the latest processes of - say - TSMC, how bad is that truly?

f. Irrespective of whether semis are indeed the new oil or not (some of the questions posed above suggest that the industry might not be as critical or unique as made out to be), perceptions are going to keep them at the center in global geopolitics. The industry makes Taiwan far more pivotal for both US and China in their rivalry. The developments are likely to make China more committed to self-sufficiency in semis. It might not be as difficult as presumed, even if they remain somewhat behind compared to the cutting edge.

g. There is a tendency amongst some US nationalists (at times in the book too) to look down upon others' progress. If they see an advanced fab in a foreign land, they feel it is because of a transfer of technology or worse, or because of the low cost of debt (never mind the low cost of equity funding in the US), subsidization, government policies, etc. A lot of this is true, but even more accurate is the rapid, unheralded, vast number of innovations - as reflected in the patents filed - in these Asian countries.

Even fifty years on, Chip War is about a fast-changing industry. It is about the world's most politically colored industry also. The book will surely become dated soon, but the sections on the early days will be helpful for any reader who wants to learn how the industry evolved in its first few phases.