Fabless: The Transformation of the Semiconductor Industry

by Daniel Nenni, Paul McLellan

The purpose of this book is to illustrate the magnificence of the fabless semiconductor ecosystem, and to give credit where credit is due. We trace the history of the semiconductor industry from both a technical and business perspective. We argue that the development of the fabless business model was a key enabler of the growth in semiconductors since the mid-1980s. Because business models, as much as the technology, are what keep us thrilled with new gadgets year after year, we focus on the evolution of the electronics business. We also invited key players in the industry to contribute chapters. These "In Their Own Words" chapters allow the heavyweights of the industry to tell their corporate history for themselves, focusing on the industry developments (both in technology and business models) that made them successful, and how they in turn drive the further evolution of the semiconductor industry.

Genres: Business, Technology, Nonfiction, Computers

Kindle Edition

First published January 1, 2013

Reviews

[Omar · Rating 3 out of 5]

★★ for most of the vendor insert sections

★★★★ for overall history and a couple of the vendor inserts

i know it isn't their job, but i'm amazed at how little vision these chip companies have. just industry analyst pablum and quantitative improvements

now that everyone has a smartphone and Moore's Law is dying…
you can smell the desperation at the end

[Moritz Mueller-Freitag · Rating 3 out of 5]

Semiconductors are the backbone of modern society, dictating everything from the processing power in your smartphone to geopolitical tussles between the great powers. They are also among the most complex technologies ever built, requiring thousands of ultra-specialized companies to collaborate across a long and intricate supply chain. It is nothing short of a miracle that this 65-year old industry can fabricate chips the size of a fingernail and containing billions of miniature transistors (each more than 10,000 times thinner than a human hair), while at the same time making them affordable enough to power everyday consumer devices.

Fabless traces the history of this remarkable industry from both a technical and business perspective. The book’s core thesis is that Moore’s Law couldn’t have been sustained without the development of the foundry business model in the mid-1980s. Until then, all semiconductor companies were integrated device manufacturers (IDMs), which means they designed and manufactured their chips in-house. This changed gradually as fabrication plants (commonly called “fabs”) got more expensive and leading-edge process technologies became increasingly complex. By the late 1980s, more and more semiconductor companies—the likes of Xilinx, Qualcomm, Broadcom, and later Nvidia—began looking for ways to outsource their manufacturing and go “fabless.”

The separation of chip design and manufacturing was finally made possible by the creation of TSMC in 1987. As the world’s first pure-play foundry, TSMC enabled fabless companies to shed the high fixed costs of running a fab and to specialize in the design of advanced chips. Much like Amazon has done with Amazon Web Services (AWS) and Fulfillment by Amazon (FBA), TSMC has allowed an entire industry to transform its capital expenditures into a variable cost. In the process, TSMC became one of the most valuable companies in the world and a de-facto monopoly at the leading edge. In the past twenty years, we have gone from 22 companies manufacturing chips at the cutting edge to just two. But considering that Samsung is now likely to have fallen behind one process generation, TSMC is the last man standing. To quote its founder Morris Chang: “The semiconductor business is like a treadmill that speeds up all the time. If you can’t keep up, you fall off.”

Unfortunately for the reader, Fabless has fallen off the treadmill as well since its publication in 2013. It provides little insight into the smartphone revolution of the past decade and strangely omits the most valuable chip design company in the world (Nvidia). Instead, half of the book is filled with quirky vendor inserts that are poorly written and annoyingly repetitive. This topic has the potential for a much better book, but strangely no one has written it yet. I sincerely hope a great writer fills this market gap soon!

[Richard Zhu · Rating 3 out of 5]

After playing several dozen hours of Factorio (look it up), I've really started to appreciate the value of electronic chips. In the game, they power all sorts of advanced technologies, like automation, upgrades, and space tech. In the real world, they underlie every new technology developed in the last several decades. The vast wealth captured by software tech giants today belies the fact that all of it is built on top of silicon. There's pretty much nothing else in this world that scales as nicely as semiconductors: most things have severely diminishing returns on investment. Moore's law on the other hand has continued to scale pretty much linearly with R&D investment for 5 decades. Thanks to the ingenuity of the semiconductor industry (optical lithography, chip design automation, etc) we've managed to sustain it for years -- and most of the value created by this 10^15x reduction in cost since the first transistor has accrued to consumers, not the industry itself.

This book is a decent history, but it's certainly not complete. I'd give a hard pass to most of the vendor inserts. As the industry has matured, so has the outlook -- the observations made by semiconductor industry veterans at the end of the book aren't particularly inspired, referring mainly to the "internet of things" as the core driver of change. I think there are more interesting things on the horizon, like probabilistic chips and non von-Neumann computers... but those will take a while to mature, if at all. The path dependence of history may mean that those may never be competitive for the same applications as semiconductor chips.

I think this book has interesting lessons for high-tech non-software industries. For example, biotech and drug manufacturing has very similar economics as semiconductor manufacturing. Most biologic drugs are manufactured by hijacking cells and persuading them to produce the thing we want. It requires a lot of upfront investment to setup a manufacturing plant complete with cleanroom. Compliance with FDA regulations only adds additional costs. There's a lot of shared technology in what cell lines are used, how the process works, and what sorts of lab diagnostics one needs to run. Whole cottage industries have sprung up to support different segments of this process. You can even see a mirroring trend towards "fabless" in biologics: most of the value in pharma accrues to the companies that patent a drug and take it to clinical trials, not the ones that produce it. If the trend towards standardization (of cell lines, fermentation protocols, etc) continues, I could see contract manufacturing making a lot more sense for folks. This can be true even if there remains substantial innovation in processes like new cell lines, cell-free manufacturing, and quicker change-outs.

So yeah, if you're curious about any of the above, this is probably worth a skim.

[Alessandro Piovaccari · Rating 5 out of 5]

A must read for everyone working in the seminconductors industry

Even if a little dated (written in 2014), this book contains many interesting insights. More than anything else is a great historic report of all the actual facts that have been driven the semiconductors industry from cosentino to the actual pervasive fabless model.

[Carter · Rating 3 out of 5]

This, is a rather, older book. It reflects a transition, that has created "fabless" companies like TSMC. The "demise", of Intel's fabrication advantage, has perhaps, also happened as a a result. An iinteresting overview, of a changing industry; with custom ASIC's and the end of Moore's Law, in the not so distant future- what changes witht the future bring?

[Jeffrey Hart · Rating 3 out of 5]

Mostly first-person accounts of the move toward fabless semiconductor firms in the last two decades. Highly relevant to understanding the current dependence of many U.S. firms on the Taiwanese manufacturer, TSMC. Intel is now responding to the challenge of catching up with Samsung and TSMC by creating its own foundry operations but it will be at least two years until it is up and running.

[Petr Polovodov · Rating 5 out of 5]

A very good reference for semiconductor industry history

[Jimmy · Rating 5 out of 5]

This book is a great overview of the foundry, fabless ASIC, EDA, and IP segments of the semiconductor industry. It discusses the origins and key technical as well as business model developments in each of these industry segments over the past few decades, including profiles of the companies like TSMC, Arm, and Cadence. It's a great book for those interested in semiconductor history who have some familiarity with the industry.

[Terry · Rating 5 out of 5]

A good history about the development of semiconductor industry, and good summary of the history of key semi companies. A must read to invest in the semiconductor industry.

[Igor Zurimendi · Rating 3 out of 5]

Interesting, but too much corporate self promotion.

[Matthew Gaines · Rating 4 out of 5]

Extremely dense.

[Emilio Garcia · Rating 3 out of 5]

Interesting in the introductory parts of each chapter but boring some of the “in their own words” narratives, the majority focused on praising and advertising the company.

[Milos Mirosavljevic · Rating 4 out of 5]

A bit outdated, but regardless, an interesting brief history of fabless companies in the semiconductor industry.

[Tang Jie · Rating 5 out of 5]

little out of time，but still provide a lot

[Hendrik Borginon · Rating 4 out of 5]

Good book, but more oriented at anyone with a professional interest in the industry rather than the general public. This history is interspersed with chapters written by former leaders of the various companies involved and this comes with a natural range of repetition and writing talent. It might be a bit dated these days, but is still a great reference to start with and build understanding of the broader landscape.

[Kursad Albayraktaroglu · Rating 5 out of 5]

This is an excellent book for anyone who is working on semiconductors, or who is interested in the history of this world-changing industry. While the histories of some of the better-known players (such as Intel) were well-documented in other books; this book is the most likely the only one that chronicles the histories of semiconductor and EDA companies like Xilinx, Synopsys, Mentor Graphics, and more. Nenni is a keen observer of the industry and his coverage and commentary of the development of the "fabless" business model is top-notch and very informative. Highly recommended.

[Riley Holmes · Rating 3 out of 5]

This is an interesting "insider history" of the most revolutionary industry of all time. He provides case studies of many different semiconductor firms, illustrating how the business model shifted over time as the cost of manufacturing increased exponentially. (For example: Apple and Samsung are top competitors in the smartphone market, but the're also each other's top semiconductor customers, due to the high cost of an idle production line.)
The book could be shorter, and I would have enjoyed more technical detail in the process descriptions.

[Oivind Loe · Rating 4 out of 5]

Very good introduction to the history of the semiconductor industry