Top>Opinion>System Technology that Fully Utilizes Hardware can Revive Japan's Electronics Industry

OpinionIndex

Ken Takeuchi

Ken Takeuchi [Profile]

System Technology that Fully Utilizes Hardware can Revive Japan's Electronics Industry

Ken Takeuchi
Professor, Faculty of Science and Engineering, Chuo University
Areas of Specialization: Green Nanoelectronics, LSI Design

Read in Japanese

"Japan is no longer suited for manufacturing."
"Hardware businesses can only be successfully started in countries like Korea and China, where costs are low."
"Apple is a service company."
"Music distribution and similar services, not manufacturing, will generate profit."

In recent times, hardware and manufacturing in particular have become the targets of severe criticism. We can see a significant rise in the presence of Asian businesses, not only in my own personal research field of semiconductors, but also in the production of LCD panels and televisions. The resulting harsh price wars have led to a series of cutbacks in and closures of hardware businesses in Japanese companies. Even the production of solar panels, which have marked expectations for future growth, already has its top share in the global market dominated by Chinese companies.

Elpida Memory, the only manufacturer of DRAM in Japan, went bankrupt on February 27 after facing an enormous \448 billion debt. After I posted information about Elpida's bankruptcy on Twitter (@kentakeuchi2003), there was a tremendous response, with 8,000 new follows in the past 2 months. My tweets regarding Elpida Memory were summarized in a Togetter titled Ken Takeuchi Speaks on the Cause of Elpida's Bankruptcy [Takeuchi ken shi ga kataru erupida tosan no genin], which was read by 115,000 people.

On the other hand, Apple recently acquired Anobit Technologies, an Israel-based venture business dealing in SSDs (Solid State Drives) and flash memory controllers, for approximately \40 billion. Apple, of course, is well-known for the development and sales of hardware including the iPhone, iPad, and MacBook Air, and mainly receives notice for its product design and strategic planning, as well as its iTunes music distribution and other services, rather than manufacturing. That being the case, what was the reasoning behind Apple's acquisition of a semiconductor venture business with no production facilities?

Even now, hardware such as semiconductors and materials hold a huge amount of potential for innovation. However, simple hardware is quick to be imitated by businesses in China, Korea, and Taiwan, and is easily swallowed up in the frenzy of price competition.

Therefore, the key to differentiation is system technology that can fully realize the possibilities of hardware. In the case of Anobit, although flash memory and SSDs are critical components included in products such as the iPhone and MacBook Air, Apple's ultimate goal is Anobit's system technology, which is able to harness the latent potential of the flash memory hardware.

Flash memory is a type of hardware originally developed in Japan, which stores data for mobile devices and personal computers. I was involved in the development of flash memory technology and the startup of its business in my 15 years at Toshiba and 5 years at university. My book, Work Skills for Global Competition: An Engineer's Struggle to Challenge Cutting-Edge IT [Sekai de shobu suru shigotojutsu: Saisentan IT ni idomu enjinia no gekisoki] (Gentosha Shinsho), which recounts the story of flash memory's development, was published this past January, and received widespread attention, reaching #2 on Amazon.com's bestseller ranking list.

With flash memory and SSDs, it is not just the memory hardware itself, but the system technology to control this memory that is becoming increasingly important. Flash memory is manufactured using the most advanced microfabrication technology in the world; for example, a product released in 2012 was miniaturized to a scale of 10-odd nanometers (1 nanometer = 1/100,000 millimeter).

Miniaturization offers the advantage of increased memory capacity. However, along with this, the number of electrons stored in a single unit of memory decreases, and interference with the space between other nearby memory units, as well as other effects, lead to the drawback of lower reliability. As the number of times rewrite operations are conducted to memory increases, the memory becomes more likely to fail. Flash memory from 10 years ago could be rewritten 10,000 times, but this number has fallen to only several thousand times for recent flash memory.

The Anobit company acquired by Apple specializes in a system technology that can correct errors in fragile flash memory (ECC: Error Correcting Code). Even if data stored in memory is corrupted due to a large number of rewrite operations, the corrupted data is repaired by a controller. This allows flash memory that could only be rewritten 3000 times to be used, for example, up to 6000 times instead.

Opting not to engage in actual flash memory production, which would require a considerable financial investment, Anobit instead focuses on controllers with the ability to effectively utilize flash memory. Although these controllers can enhance the quality of flash memory several to ten times over, their development does not require such significant funding.

If even just the concept for a superior error-correcting algorithm can be devised, a venture business, or even a university research laboratory, can design a controller semiconductor chip without the need for a large investment. The actual manufacturing can then be consigned to a foundry (contracted semiconductor manufacturing company) such as TSMC or UMC in Taiwan.

Currently, I am also personally involved in the research of flash memory control systems and controllers, and so Anobit is my rival. This past February, at the ISSCC (International Solid-State Circuits Conference), also known as the Semiconductor Olympics, the Takeuchi Lab unveiled a controller that can extend the lifespan of memory by 10 times, attracting the attention of both internal and external media. Just when we thought we had succeeded in surpassing Anobit with this technology, we heard of Anobit's acquisition by Apple.

Many of the engineers at Anobit were originally from Msystems, a controller manufacturer also in Israel, and while I was working at Toshiba, I collaborated with Msystems in flash memory controller development. Hearing of friends and acquaintances acquiring billions, or even tens of billions, worth of yen in wealth not only surprised me but frankly, gave rise to a range of mixed emotions. I have some personal regrets that I didn't also start up a venture business, and I feel that perhaps my turn will be coming up next.

Just as with the systemization of hardware described above, in the electronics industry, Management of Technology (MOT), which is a link between technology and business, is becoming more and more important rather than simple technological development. I have a running column on the Nikkei BP homepage, MOT that Engineers Should Know, by Ken Takeuchi [Takeuchi ken no enjinia ga shitte okitai MOT], where I provide instruction and education in MOT.

Although the outlook of Japan's electronics industry may be grim at the moment, it is too early to give up. There are still great possibilities for the hardware technology that is Japan's strength. Through implementation of MOT education and the development of system technology that can fully utilize hardware, I strongly wish to play a part in the revival of the electronics businesses of Japan.

Ken Takeuchi
Professor, Faculty of Science and Engineering, Chuo University
Areas of Specialization: Green Nanoelectronics, LSI Design
Born in Tokyo in 1967.
Graduated from the Department of Applied Physics, Faculty of Engineering, The University of Tokyo in 1991.
Completed the Master's Course in Applied Physics, Faculty of Engineering, The University of Tokyo in 1993.
Completed the Master of Business Administration (MBA) program at Stanford University in 2003.
Acquired the PhD degree in Electronic Engineering Course, Graduate school of Engineering, The University of Tokyo.
Professor of Engineering (The University of Tokyo).
Following positions at Toshiba and as Associate Professor, Department of Electrical Engineering and Information Systems, Graduate School of Engineering at The University of Tokyo, he assumed his current position in 2012. His current research is in extremely low-power, high-capacity memory and computer systems. He provides education in Management of Technology (MOT), striving for the goal of developing T-shaped persons, with a wide-ranging perspective and also deep specialized knowledge, mirroring the horizontal and vertical legs of the letter T. He has succeeded in the practical realization of flash memory at Toshiba, has developed products with world-leading memory capacities a total of 6 times, and also holds 210 patents worldwide.