In the fast-moving semiconductor world, it’s kind of peculiar how the same players develop similar products but they rarely manage to make the one truly great hit. Ignore the software developers who are off on cloud nine with their abominable “social networking” software packages. I’m talking about hard-core products and associated programs that real engineers put together.
Having said that, does the name “Nakamura” mean anything to you? Probably not, unless you are a Japanese semiconductor designer. In Japan the name is especially well known in the rapidly expanding world of electric lamps that are not incandescent. LEDs and LCDs are prime examples of this new breed of industrial and domestic lamps. Professor Shuji Nakamura has had a hand in the development of all these non-incandescent lamps.
Nakamura’s latest and brightest is an LED based on gallium nitride (GaN). The problem is that GaN chips are difficult to manufacture in large quantities. This is where Professor Nakamura steps in. In a prior incarnation, Nakamura successfully developed the blue semiconductor LED, a notoriously difficult thing to do at the time.
Just to make life even more difficult for him, Nakamura’s then employer, Nichia, threw the Japanese legal code at Nakamura, who in a Dickensian battle won his case but who also lost his prime research position at Nichia. Fortunately, Nakamura found a home in the well known Materials Department at UCSB (University of California–Santa Barbara). He also joined a team of developers of GaN devices under the wing of Eric Kim, a former Intel executive. Kim’s company, Soraa Inc. (Fremont, CA), recently obtained more than $100 million of venture financing with the help of legendary Silicon Valley venture capitalist Vinod Khosla.
Nakamura also won the Finnish 2006 Millennium Technology Prize, worth about $1.5 million, for his work on blue and white LEDs. This is the technology used today in the high-performance Blu-ray disc players.
Soraa is facing several major international makers of incandescent lamps, many of which are also researching LED lamps. The makers are working hard to get the manufacturing costs down for the market-leading 60-W-equivalent lamps. The Lumileds division of European giant Philips NV expects to have a single-chip lamp comparable to the conventional 50 W bulb for a retail price of $50 by year’s end.
Soraa is expected to have a comparable product available using a technology called GaN on GaN. This product can be made on a conventional silicon wafer line, thus bringing down manufacturing costs. However, the yield of a GaN wafer is much lower than a conventional silicon wafer line. Nonetheless, Nakamura is confident that Soraa will have a competitive product by year-end.
Soraa may be in the right place at the right time with this new product. According to Ryan Sanderson of IMS Research, “Demand for LED lighting solutions is increasing rapidly for all applications from low-power residential retrofit LED lamps to high-power commercial and industrial LED luminaires for applications such as street lighting.”
Although Soraa appears to have the pole position with GaN on GaN technology, several Taiwanese semiconductor makers are reporting significant success with alternative technologies, such as GaN on silicon. In my view, the companies with the deepest pockets for refining their manufacturing processes will win the initial battles, but in the long run, single-chip lamps appear to be poised to win overall.
Once more, Shuji Nakamura seems to be leading the pack in research, but Soraa may not have the long-term funding of a Phillips or a GE to become a leader in the mass-market production of LED lamps.