The book of the blog is now available for purchase here. Here’s an extract from the introduction that gives an overview of what the book is:
This book is an outgrowth from this blog. Although the basis of the book is the original blog entries there is new material and the old material has been extensively revised. Furthermore, I’ve reordered the content so that each chapter covers a different area: Semiconductor, EDA marketing, Investment and so on.
I’ve tried to write the book that I wish I’d been able to read when I started out. Of course it would have been impossible to write back then, but in a real sense this book contains 25 years of experience in semiconductor, EDA and embedded software.
Not many books cover as wide a spectrum of topics as this. I’ve ended up in marketing and executive management but I started out my career in engineering. For a marketing guy I’m very technical. For a technical guy I’m very strong on the business and marketing side. It is an unusual combination. I started in semiconductor, worked in EDA but also in embedded software. I’m a software guy by background but spent enough time in a semiconductor company that I have silicon in my veins.
I certainly won’t claim that this is the only book you need to read, so the end of each chapter has a “bookend” that looks at one or two books that are essential reading for that area, a taster to encourage you to read the book yourself.
There are a few threads that recur through the whole book, leitmotifs that crop up again and again in widely different areas.
- Moore’s Law, of course. But really something deeper which is that semiconductor economics, the fact that transistors have been constantly getting cheaper, is the key to understanding the industry.
- Software differentiation. The reality is that most of the differentiation in most electronic systems is in the software (think iPhone).
- FPGA-based systems. Combining these two trends together, chips are getting too expensive to design for most markets so “almost all” designs are actually software plus FPGA or standard products.
- Power is the limiter in many designs today. We will be able to design big systems but not be able to power up the whole chip at once.
- Multicore hits all these trends. Nobody really knows how to write software for multicore designs with large numbers of cores, nor how to handle the power issues.
- Finally, all the major breakthrough innovations in EDA have come from startup companies, not the big EDA companies. But the exit valuations are so low that no investment is going into EDA startups so it is unclear where the innovation will come from going forward.
The book is available for purchase here. If you are interested in buying in bulk (makes a great giveaway for your best customers at DAC, or for your user-group meeting, or for your sales kickoff) then contact me directly. I can also put a special cover on it with a company logo.