2005 Outlook: Semiconductors –

If you opened one of Pierre Lamond’s veins, it would probably bleed silicon. He’s been involved in the semiconductor business since its birth in the late 1950s, when he was general manager of Fairchild Semiconductor’s Integrated Circuit Division. He went on to co-found National Semiconductor, where he was Technical Director. After 30 years on the management side, he joined Sequoia Capital in 1981, where he continued to build new chip companies, as well as software and networking startups. His venture resume includes such household names as Cypress Semiconductor, Redback Networks and Vitesse Semiconductor.

Q What is your outlook for 2005?

A The venture industry invests for five years from now, so it doesn’t care so much about 2005, but the semiconductor industry turns on near-term cycles. As VCs at Sequoia Capital, we’re aggressively pursuing our investment strategy.

What do VCs see as their strategy for 2005?

First, as VCs we find it more difficult to find the exceptional deals. Today, everything revolves around architecture and design. This has been true for years. Process technology has been in the hands of the foundries, TSMC or UMC, so VCs have to invest in innovative designs. There has been no investment in a new IDM for several years.

In sectors surrounding semiconductors?

Yes, there are opportunities in areas like cooling and MEMS. We’re in one or two of these. But EDA software, developing tools for 90- and 65-nanometer generation designs is really tough. The field is dominated by four firms: Synopsis, Cadence, Mentor and Magma.

The materials area?

Yes, materials like indium phosphide, where there are applications for high-performance amplifiers for cellular applications. But there are already 20 small indium foundries that were started at the peak of the bubble.

Is there a change in the leadership of the semiconductor industry?

No, I think Intel is still the leader, but there are challengers such as Samsung. Semiconductors are not a monolithic industry just driven by the microprocessor. There are products like micro-controller, reconfigurable logic, ASSPs and ASICs. All are in different sectors, walking at different rhythms.

Do VCs still have the opportunity to found the next Intel or Xilinx?

I think it’s possible. Look at automobiles, which were dominated in America for generations by three firms. Then the Japanese manufacturers entered the field and now there are many firms that lead automobile manufacturing. The next Intel may already be here and we don’t know it yet.

If selection is the key to VC success in semis where do you invest?

Well, we don’t like to talk about our strategy at Sequoia, but I can tell you what we haven’t done: We haven’t made a single investment in standards-based semiconductor design areas like 802.11. It’s too hard to figure out who will be a winner in such areas, and even if you succeed the returns aren’t very great. We try to invest in areas where we can differentiate ourselves.