1006EN-Semi
Keywords: clusters, Oregon
@head: Oregon Can Still Have A High-Tech Future
@deck: The “white spaces” of converging markets can be filled in for future technological growth.
@text:The talk started simply enough. First, there was the opening joke: Oregon’s SEH produces more wafers per year than Vatican City. Murmured laughter was followed by the gentle clinking of dessert forks, which were being raised for the last time at the Seventh Annual SEMI Pacific NW Outlook Dinner ( Keynote speaker David Y. Chen, a partner at OVP Venture Partners, got straight to the point: What was the growth potential for Oregon’s high-tech community? He began by noting that Oregon represents the fifth largest technology region in the U.S. Yet the state was still recovering from the technology depression in 2000, which hit the entire country. Although Chen didn’t mention the continuing Intel layoffs, I’m sure that thought was on the minds of many attendees. After all, Intel maintains a strong presence in the Portland area.
With this lead-in, Chen went straight to the question at hand: Where does a region like Oregon--indeed, the Pacific Northwest as a whole--grow next? His response was to look for the “white spaces”--that is, the areas of converging technological markets. For Oregon’s semiconductor industry, Chen concluded that these white spaces can be found in four technology areas: multi-core and programmable devices; nanotechnology; wave-energy systems and the design-construction of sustainable architectures; and the integration of life sciences with the semiconductor community.
As I listened to Chen’s talk, I was reminded of our pre-dinner interview. I had specifically asked him about two of Portland’s emerging and--in my opinion--converging technology clusters: software development and programmable devices. The newer multiprocessor programmable fabrics, such as Oregon startups Ambric ( and Mathstar ( are necessitating a corresponding scaling of existing software at the real-time-operating-system (RTOS) and application level. In other words, software must scale upward to support emerging multiprocessor devices. A single chip that contains eight processor cores with four threads per core, for example, will appear to software as a 32-processor system. To take full advantage of such a chip, software will need to scale upward in terms of development efficiency—that is, unless application development companies intend to hire 32 separate software programmers. <grin>
This trend toward next-generation, multiprocessor systems-on-a-chip (SoCs) was highlighted by Daya Nadamuni, former Research VP at Gartner Dataquest, during this year’s Design Automation Conference (DAC). Second-generation SoCs are single-chip, multi-functional (that is, multiprocessor) devices. Each of these chips drives a subsystem with its own operating system, firmware, and software application-programming interface (API). Examples of second-generation SoCs include Texas Instruments’ OMAP and Philips’ Nexperia, which are used in today’s mobile smart phones and video cameras. [See this issue’s DotOrg section for more about Philips’ internal SoC tool, which was used to develop the Nexperia multimedia-IC product line.]
Chen agreed that the Portland area is well posed to succeed in the emerging multi-processor, second-generation SoC market. Oregon can support this market through two of its several technology clusters: software development and programmable chips. Portland is already well known for its open-source software community ( It also was one of the first areas to offer a graduate-level software-engineering degree through its combined university program. As mentioned, the Portland area’s programmable-device cluster consists of such startups as Ambric ( and MathStar ( Ambric is especially focused on software-scaling issues for multiprocessor systems.
What factors are needed to grow technology clusters in software development, programmable devices, and other areas? After Chen’s keynote address, this question was asked of a panel of experts. Their responses centered on the key elements for technology growth: IP, capital, infrastructure, and communication. Many experts believe that technology clusters form around a center of intellectual property like universities and labs. Capital follows this IP. But capital investments won’t stay around if an infrastructure--for both data and personnel--is missing. The final ingredient for a successful technology cluster is communication--that is, courting media to spread the word.
A common thread among the panelists’ responses was the need to realistically fund engineering and science programs. One panelist noted that China calls electronics a “pillar” industry, which is in stark contrast to the meager visibility allotted to the electronics industry in the U.S. Another member suggested that attendees write their senators to demand funding for science and engineering, as these professions drive our economy. With only seven engineers in the House and Senate and five scientists in the House (according to Congressional Quarterly), I doubt that such pleas will be appreciated or acted upon. Plus, all of these good ideas aren’t enough without inspiration, as Chen noted in his closing remarks. Americans need to rally around some sputnik-like cause--perhaps energy independence. While this idea isn’t new, it does seem to offer hope that Oregon and the nation will continue to have a future in the technological world.