This session included three presentations: Present & Future Of Digital Technology by JJ (Junshi) Yamaguchi, executive vice president, NEC Electronics; Mobile Device Technical Challenges by Petri Liuha, laboratory director Computation Structures, Nokia Research Center; and The Other Barrier In Electronics: Applications by Kees Van Der Klauw, senior VP of Technology & Development, Philips Consumer Lifestyle.
Yamaguchi opened his talk on the Present & Future Of Digital Technology by pointing out that ubiquitous computing is on the horizon. In other words, society is transforming to a time when computing is accessible to anyone—at any time and in any place. He supported this statement by showing slides on the explosive growth of mobile phones, especially in emerging economies, with faster communications on these phones enabling all sorts of computing experiences (see Fig. 1). He also pointed to rapidly falling PC prices as a driver for ubiquitous computing and showed the $300 ASUSTek Eee PC and $200 computers offered by the One Laptop per Child (OLPC) project as two examples of this trend.
Yamaguchi touched on many areas where semiconductors play an important role including mobile devices, flat panel TVs, home networks, hybrid cars, automotive entertainment, and medical devices. He concluded that the semiconductor is one of the key devices needed to achieve ubiquitous computing via connectivity, affordability, and attention to the environment.
Liuha’s presentation on Mobile Device Technical Challenges showed how the complexity of mobile phones has increased over time, so that today we have a multimedia computer in our hands rather than just a mobile phone. But as complex as today’s mobile devices might be, there is still the need for increased power efficiency. He also pointed to the changing nature of business models from a telco model where access was key to an Internet model where content is most important. As an example, he showed a “point and find” application on a cell phone that allows a user to simply point their phone at an object to access information about that object (see Fig. 2). He summed up the challenges by dividing them into several categories: the radio, computing, interoperability, sensing, and rendering of media. On the radio front, he suggested that a handset would need as many as 10 antennas that could take up 25% of the handset’s volume and would need multiple air interfaces. For computing, a handset would need the power of a notebook with 50 times the power efficiency. Finally, he thought a handset would have to interact with sensor networks and recognize physical objects as well as handle a variety of media formats.
Van Der Klauw’s talk on The Other Barrier In Electronics: Applications began with a look at Moore’s law and how it has continued to be a useful metric over time. He made an interesting observation regarding Moore’s law. To keep on track over time, a snap-back effect and a stacking of generation curves (see Fig. 3) occurs over time due to breakthroughs in technology. But with each new generation there are diminishing returns due to increasing investments and a value erosion in applications. He made the point that the process industry is in conflict with the consumer electronics industry. Whereas the process industry follows Moore’s law and eventually leads to commoditization, the CE industry thrives on differentiation. The application barrier as Van Der Klauw sees it is embodied in Figure 4, which shows the graphs of successive process generations bounded by a commodity “roof” and a returns “wall.” Lifting the roof up enables value creation, while pushing the wall to the right leverages investments. In one example of how this might be accomplished, Van Der Klauw showed a commodity SoC on one slide, and then a differentiating SoC on another (see Fig. 5). As shown in the figure, adding a switch matrix and/or sea of gates to the SoC enables differentiation.
To pass the applications barrier, Van Der Klauw concluded that more open architectures are required in industry, systems, hardware and software in order to leverage Moore’s law while enabling differentiation and value creation
Reprints
