The Chip Industry Is No Longer What It Used to Be

The chip industry these days feels sort of like that noteworthy transition when e-mail was discovered as the killer Internet application. No longer were the telephone or fax machines used as critical tools for differentiated, cutting-edge communications.
 
Likewise, the chip industry, in so many ways, is no longer the way it used to be just a few short years ago. For example, I'm starting to wonder whether we are really just a chip industry anymore, with so many of our concerns focused on systems, software, electronic design automation tools, and inventory turns.
 
I talk about systems and software more often, it seems, than I do chips. In the network processor chip market, for instance, more than half of the buying decision comes down to which company offers better software.

Like software, much of the time the industry seems preoccupied with chip tools rather than the chips themselves. Differentiation using chip tools is becoming more important than ever.

In the telecom chip arena, it's no longer about making a bunch of different types of chips for a bunch of different types of telecom equipment; it's about making a one-size-fits-all chip that works in many types of equipment. It's no longer so much a question about the price of a chip but rather what the total system costs, such as PCB, chip, and design layout expenses; chip usage; and what and how chip tools are being used.

In the home-networking chip arena, it's no longer about making modem chips that enable 56 Kb/sec Internet connections between two home computers. It's about combining gigabit Ethernet speeds with storage technology onto a single system chip to create the network in a single box, rather than a bunch of different types of boxes spread all over the house. It's about putting a network on a chip, essentially, rather than several different chips on a circuit board that form a network.

In the gigabit Ethernet chip arena, it's no longer about making chips that pump these speeds around big offices; it's about making these chips capable of being used in small offices and homes. It's about the applications of the chip more so than the chip itself.

So what does this trend regarding the chip industry's distractions away from chips mean to you and me? First of all, it doesn't mean we need to get out of the chip business because it remains a robust industry and will for many years. There are many ways for companies to continue to make the most integrated, cost-effective, highest-performing chips they can.

But it does mean we need to think more carefully about what kind of companies we used to be, are today, and want to be tomorrow. If a chip company believes its differentiator is software that runs on its chips, it should invest more time and resources improving on and leveraging to help those differentiators. If a chip company believes the way it uses tools differentiates it from competitors, it should highlight and build on that strength.

Making a high-performance chip is really only one of many factors that determine whether a company succeeds or not. Superior customer service, for example, can be a chip company's key differentiator in a market where the products are very similar. What reigns as supremely important, and will continue to, is how fast a chip company can deliver a product to market.

If a company has market-leading functions on one chip that outperform alternatives, it should find ways to leverage those functions on other chips.

Yes, the chip industry's evolution during the past few years has taken us to a place where it's becoming tougher to isolate a company as a "pure" chip company. There's so much more going on than ever before in this industry in terms of sophisticated associated technologies, changing market dynamics, and radical strategies to seek competitive differentiation.

The question for you to ponder is this: Which companies will embrace this concept the fastest — that they're not just chip companies anymore and can't afford to be — and make moves to ensure they retain differentiation in other ways?

Which companies want to remain "pure" chip companies, and which ones don't? Which ones want to become leaders in chip software? Which companies want to derive competitive differentiation from systems knowledge of shrewd applications of chip tools?

It's no longer an industry that is as simple as a piece of silicon on a circuit board.