The Journal The Authority on Global Business in Japan

Welcome to the era of the merged identity, the industry powerhouse. I’m talking about the semiconductor industry, which is in transition as we move to a smarter world where everything is connected to a network.

Activity in the industry has stepped up a gear in recent months, as the leading players have set out their game plans for the future and have commenced bulking up for the fight ahead. Europe-based market-leader NXP Semiconductors bought US rival Freescale Semiconductor Inc., creating a $40 billion enterprise, the Financial Times reported in March.

And as of April this year, there was speculation over Intel Corp.’s plans to launch a shopping spree in the industry—including talk about the multinational corporation buying Altera Corporation, a leading integrated circuit maker, in a $10 billion deal.

Meanwhile, others have conjectured that Intel would buy fellow chip supplier Broadcom Corporation, based in Irvine, California, for what could be—at $30 billion—a much bigger deal.

The industry wants to realize opportunities for lots more semiconductor use in our everyday lives, from driverless cars to domestic robots to ingestible micro-devices (computerized pills) that can monitor and produce personalized health data.

And it’s fair to say the winners in this market will generally have one thing in common: a merged identity.

“It’s the combined company that will capitalize on the growing opportunities created by accelerating demand for security, connectivity, and processing,” says Hillary Cain, director of marketing and communications, Americas, at NXP Semiconductors.

The past is another world
The semiconductor industry—upon which the microchips of modern electronics devices rely—has changed radically since it really got started back in the ’80s. The figures say a lot.

The global market in semiconductors was worth around $20 billion in the second half of the ’80s, rising to around $50 billion as we moved into the ’90s.

Jump forward to 2015, and we see the market set to be worth some $344.5 billion worldwide, according to World Semiconductor Trade Statistics, an industry-led organization that provides market data and forecasts. And yet Japanese semiconductor companies seem to have slipped back during this period of plenty.

One person who remembers what it was like in the ’80s is Raman Chitkara, global technology specialist at PricewaterhouseCoopers, a multinational professional services company.

“I remember back then, you couldn’t go to a business meeting in Silicon Valley without someone talking about the rising domination of Japanese semiconductor companies,” he says.

“In 1985, out of the top 10 semiconductor businesses, five were Japanese, four were US-based, and one was European. Then, in the ’90s, out of the top 10, six were from Japan!

Now I’m looking at 2013, and I can see that out of the top 10, just one Japanese company remains. Of the others, five are US-based, one is European, two are South Korean and one is Taiwanese,” he adds.

The arrival of fabless
This begs the question: What went wrong with the Japanese semiconductor industry? Chitkara has an answer.

“Look at the fundamentals of this market and what has changed in the last 25 years. We’ve seen the rise of the fabless model, where manufacturing is outsourced [typically to semiconductor fabrication or “fab” companies in countries such as Taiwan]. And we’ve seen start-ups playing a big role in driving [fabless] innovation.”

He points out that Japan neither embraced the fabless model nor promoted its start-up culture. Its domination, Chitkara says, sowed the seeds of its future failure.

“Japan held on to the old models too long. Eventually, they started a process of dis-integration, but it was 10 years too late.”

The Japanese semiconductor industry reached a low point when Tokyo-based Elpida Memory, Inc., once the world’s third-largest maker of PC memory chips, filed for bankruptcy in 2012.

With $5.6 billion in debt, as reported in the New York Times in February 2012, the company had little choice—it seemed a perfect storm put paid to its fortunes.

Elpida Memory earned a lot of its revenue from dynamic random access memory (DRAM) chips—the most common form of memory found in PCs, laptops and workstations.

The trouble for Elpida Memory was a combination of a worldwide glut of DRAM chips—which forced unit prices down—and tougher competition from firms such as Samsung in South Korea.

Add to this the strength of the yen at the time and Elpida Memory had major problems on its hands. As Yukio Sakamoto, president of Elpida Memory, suggested in an interview with the Nikkei Asia Review in 2013, the bankruptcy proceedings were the result of bad timing and ill luck.

He said the smartphone market, which was then about to take off, came too late to change the fortunes of the DRAM market and Elpida Memory.

Despite the challenges, others eyed Elpida Memory favorably, and in the year it filed for bankruptcy, US-based Micron Technology Inc. acquired the business. Micron’s Chief Executive Officer Mark Durcan said at the time, “We are creating the industry-leading pure-play memory company.”

The acquisition gave an important boost to Micron, which automatically became a much bigger player in the market, with access to all of Elpida Memory’s customers.

Sakamoto was proved right because, as soon as the mobile-device revolution kicked in, the DRAM glut disappeared and demand for production increased. The market, moreover, has remained strong.

Indeed, when Micron announced its second-quarter earnings for fiscal 2015 (ending in August) in April, it reported that demand for DRAM continued to increase even though it expected its own DRAM production to come in below the market, as reported by Forbes in April.

The future of connectivity
Eleanor Roosevelt, former first lady of the US, once said, “I never waste time looking back.” And maybe that’s the message the Japanese semiconductor industry should focus on right now: look ahead, for the industry potentially has a very bright future.

With the Internet of Things—where hardware and software are combined in a myriad of ways—becoming a reality, new opportunities are opening up. Industries such as healthcare, security, and automobiles are incorporating more semiconductor technology than ever before, and the trend is set to increase.

For Chitkara, Japan should be particularly excited about developments in transport.

“Just think about the expertise that Japan has in building automobiles. The experience of its engineers is one of the country’s key strengths,” he says.

Applying this knowledge to connected cars—which are networked to other devices via the Internet—could give Japan an important edge. Still, it could take alliances or mergers on the scale of that between NXP Semiconductors and Freescale Semiconductor Inc.—the latter company having strengthened the former’s position in the automobile market—to help Japan achieve its potential.

No doubt conversations are being had in the industry in Japan and across the world about alliances and acquisitions as semiconductor companies think more closely about their strategies.

The deals may be between chip companies or, as Chitkara points out, between a chip company and a fabless foundry. What seems clear is that the Japanese semiconductor industry has a lot in its favor, with plenty of industry powerhouse potential.

And while its dominant position of the ’80s and ’90s won’t be returning any day soon, it can again become a powerful player in semiconductor industries worldwide.

RichardJolleyBio

Divider

Richard Jolley is an IT and business writer in Tokyo.

Divider

And yet Japanese semiconductor companies seem to have slipped back during this period of plenty.