Electronics Weekly Electronics Design & Components Tech News
The Japanese semiconductor industry is facing a huge challenge. From 51% market share in the late 1980s, Japan’s market share has fallen 10% every decade since – to 41% in 1994, to 29% in 1998 and to 20% in 2010, points out one of the industry’s brightest thinkers, J. J.Yamaguchi of Renesas.
The Japanese strategy has to be one of ‘selection and concentration’, says Yamaguchi, and a focus on the new semiconductor industry emerging around environmental products, healthcare products and safety/security products tied into the high-speed communications ICs required to operate in the Cloud.
In practice this means looking at facilitating progress in mobile terminals (phones, tablets, mobile PCs, communications (LTE, HSPA+, Wimax etc), in servers/storage and in sensors, says Yamaguchi.
As people in emerging countries get richer there’s also increasing demand for what Yamaguchi calls ‘preferential consumption products’ – cars, phones cameras etc.
The trends Yamaguchi identifies are: the profitability of consumer products falls as their scale grows; business models combining hardware and software are growing; new growth areas are emerging from social needs; the share accounted for by merging markets is expanding; the cost of investment in plant and equipment continues to grow.
He sees increasingly stiff competition in ASSPs and a trend for companies to make fewer products.
Yamaguchi’s advice to the Japanese chip industry is to use Japan’s high standing in the industry to concentrate on high market share products like NAND, MCU, DRAM. CMOS sensors, MCUs and power devices, and exploit the strengths of the IDM such as high reliability, small quantity and large volume manufacturing.
Most startup action seems to be more in IC processes than in digital ICs themselves. There are some very interesting ones around at the moment both in the US and Europe.
There are also a few EDA ones but harder to get funding for them as whereas the business plan used to say “sell to highest bidder of Cadence/Mentor/Synopsys”, EDA is now so segmented that there is only usually one purchaser for any EDA technology.
Well there is one mixed signal fabless start-up (well now about 10 years old I think) in Japan, called THine, Arun, but apart from a flash start-up those are the only two start-up fabless companies in the country, so far as I’m aware. I don’t think I can get down to Si 2011 but I aslo noted the changed status of Simon Knowles. As you say, he, or anyone else, is going to find it tricky to get funding for a digital IC start-up. But maybe he feels that he’s already earned the entrepreneur title by being on the founding teams of Element 14 and Icera – unless, of course, to be judged a real entrpreneur you have to have an IPO under your belt.
That’s true but sadly the same is somewhat true in Europe and the USA. There are very few new digital start-ups and I’m honestly not optimistic about most of them. Mixed-signal is more interesting but I don’t know enough about the dynamics in Japan to say whether there’s potential there for them.
Are you attending Silicon 2011 in Bristol? I’m not but I’d be quite curious to see what the sentiment is there. Also I notice Simon Knowles has now left NVIDIA and his newly attributed ‘silicon enterpreneur’ title seems to imply he doesn’t plan to retire, hmm!
Yes, Arun, but it’s fabless in the wrong direction i.e. IDMs going fabless rather than engineers starting up new fabless companies.
Well fabless is finally happening in Japan for the leading-edge nodes. The big question in my mind is whether they will manage to penetrate international OEMs. Renesas Mobile needs will need to add a lot of western clients to be profitable. They are competing against giants so they can’t reduce their size further without being left behind on the technical side.
You’re absolutely right about factories, Greg, but it’s the rise of the fabless model which has kept the West in the game and, in Japan, fabless hasn’t happened.
1970’s America? Its extremely difficult to maintain an industrial base when free markets are always looking for the cheapest factory slave. It’s all about China.