Partly Sunny, With A Chance For Explosive Growth

I recently attended a session at the Mentor Graphics User Conference (User2User) in San Jose that dealt with the changing foundry landscape. The session was moderated by SemiWiki’s Dan Nenni and included:

• Giorgio Cesana, director of technology at STMicroelectronics
• Jack Harding, co-founder, president & CEO of eSilicon
• Lluis Paris, deputy director of worldwide IP alliances at TSMC
• Wally Rhines, chairman and CEO of Mentor Graphics

The IDM, ASIC, foundry and EDA perspectives were all represented well by these gentlemen. A few things are noteworthy about the event beyond a balanced point of view, however. First, attendance at User2User was quite brisk – the huge room was full. Comments were made by Dan Nenni (who goes to conferences for a living) that attendance is up at all events of late. Why is that?

The group of panelists discussed this question in some detail and I think a punchline emerged that is worth pausing on for a moment. There were two parts to this discussion:

1. Foundry technology is exploding. We’re not on a one-dimensional Moore’s Law trajectory anymore. Of course everyone is watching 14/16nm, and then 10nm, etc. But now, we also have FD-SOI to consider, as well as highly optimized 28nm flavors, as well as internet of things (IoT) optimized technologies in older nodes, as well as 2.5D interposers, as well as 3D stacking. You get the idea.
2. For the first time in a long time, there is an increase in startup activity with a thirst for semiconductor technology. Yes, most VCs have stopped writing checks for fabless semiconductor startups. But IoT startups that leverage semicondutor technology – that’s a different story. Jack Harding commented that he’s now seeing prospects contacting eSIlicon with new and novel ideas on a regular basis. It’s been a while since that was commonplace.

Does the confluence of broadening technology options and new semiconductor applications give rise to a bright forecast for the future? Is this why conference attendance is up? Are we on the edge of a new era of technology growth for semis? The panel largely felt this was the case.

This next phase of growth will have some important differences however. These differences were also discussed by the panel, and several questions from the audience fueled the discussion, as well. The central difference has to do with how silicon demand will be fulfilled. We will still have the “big, bad chips,” designed either by fabless chip companies or, in increasing frequency, by forward-looking system companies. But a growing segment of the user base will be product companies with a thirst for custom silicon without a lot of knowledge, time or money to to get the chip specified, designed and built.

So the questions is, how is this new type of demand serviced? How do we make semiconductor device specification, design and fulfillment more transparent and simpler? Let’s examine how other industries have addressed these challenges.

Does online, self-sevice procurement with real-time feedback sound familiar? We’ve all experienced these benefits in other endeavors, both business-to-business as well as business-to-consumer. These same benefits are coming for the semiconductor industry. eSilicon started bringing semiconductor design and manufacturing to the internet a long time ago with our eSIlicon Access work-in-process tracking system. Recently, we began to add online quoting and IP browsing. There is a lot more coming. Others will follow as well. I think the forecast is bright. Watch this space carefully in the coming months at www.silicon.com.