Align your core strengths, manufacturing, product portfolio, message, and targeting carefully
It is a very dangerous error to fail to align the chain of getting products to customers. Yet, it is a very common error. Most companies, when they do make it, fail to recognize it as the source of their problems. VLSI Technology was the first high profile company to enter the ASIC market. They essentially created it. Yet their first fab was designed to do what its management knew best: commodity DRAMs. This created a responsiveness problem that cost them their lead. Companies that get it right, sail to leadership positions – even start-ups. Back in the eighties, Novellus broke out because it was one of the first equipment companies to outsource its manufacturing. More importantly, it built its business model around this manufacturing strategy. There were lots of companies entering CVD in those days. Novellus won. Trillium jumped ahead because they broke convention in test, when everyone was trying to build the first VLSI testers conventionally. The convention was to use ECL. Trillium jumped on the CMOS ASIC bandwagon that was sweeping the semiconductor industry. CMOS brought many benefits: greater integration scale, lower power, no need for water cooling. It made for a simpler, lower cost tester that also benefited the customer. Jim Healy aligned this to their marketing. Instead of selling it as another VLSI tester, the message centered on all the operational benefits the system had. This set it apart from all the other ECL-based VLSI testers. Today, CMOS is the standard for all testers. But using ASICs has gotten to be increasingly expensive. Big Iron testers generate lots of heat and now need to be water cooled. Designing 90nm ASICs is exorbitantly expensive. It is hard to spend as much as $10M per design for a part that you only need a few hundred of a year at most. The test equipment start-ups Inovys and Nextest also succeeded by aligning their product design with their markets. They used off-the-shelf FPGAs and DIMM memory modules. While the testers couldn’t operate at high frequency, they do not need to hit their target markets – hence, their success.
This maxim is not just for start-ups. When Intel faced disaster in the mid-eighties as the Japanese onslaught was crushing American chip makers, they made the acclaimed decision to get out of memories. Others did the same, but failed. Andy Grove orchestrated the strategy and as he puts it, “We organized the company to support the strategy, so we had a coherence of market, business strategy, and manufacturing and technology strategy. When these are aligned up, things go pretty well.”