In last month's column, I profiled field-programmable analog array startup Anadigm. This month, I turn to Peregrine Semiconductor. I love it when startups take calculated risks, and Peregrine is taking two: pioneering the commercialization of a substrate technology and owning a fab.
Founded in January 1990, Peregrine spent the first seven years of its life developing a commercially viable silicon-on-sapphire (SoS) process technology. Synthetic sapphire, with chemical composition of Al2O3, functions as an insulating substrate. To this substrate, the company adds an epitaxial silicon layer upon which traditional CMOS devices can be fabricated. The main benefit of an insulating substrate is isolation that in turn enables integration of both very high-frequency devices as well as passives.
With its manufacturing technology mature, Peregrine has shifted the focus of its 125 workers and $95 million (so far) to designing devices for target markets that benefit from its substrate technology. Though sapphire substrates are not cheap-$350 for 6-inch wafers in volume-high-frequency operation and rad hardness cater well to the mil arena.
Generally, semiconductor startups with novel technology platforms have two options. One, they can create devices that uniquely serve an emerging market. Peregrine tried that in the past by marrying the transparent sapphire substrate with CMOS integration for unique parallel optical module subassemblies. Or, they can leverage their technology to better serve an existing market in terms of price-performance. This is the course Peregrine is on with a strategy of developing more highly integrated components for cellular handset antennae switch modules.
And did I mention Peregrine owns a 3,000-wafer/month fab? Though the company uses standard CMOS processing once the substrate is prepared, Peregrine felt it needed the fab to accelerate technology development and tweak device performance.
Peregrine is now entering its most critical phase as a startup. With an above-average cash burn due to fab costs, the outcome of the company's current product development will spell glory or doom. If Peregrine succeeds, it has a shot at becoming the next Cree-a company with a similar strategy based on dominating silicon carbide process technology.
Jeremey Donovan (jeremey.donovan@gartner.com) is chief analyst at Gartner Dataquest.
http://www.eet.com