Date: Friday , February 02, 2007
Pratap Reddy, Chairman & CEO, ArchPro Design Automation
With the proliferation of electronic gadgets, OEMs are making sure that they offer the best of the features to consumers. More applications packed into the small mobile device only means a big strain on the battery life. So going forward, more than the feature sets power will become the differentiating factor. Power will have a profound economic impact. At the same time with the changing geometries of design process, IC and SoC designers are facing low-power design challenges.
With power becoming the biggest component of the overall economics of a chip, I believe there will be a complete overhaul of the EDA industry. For some 20 years, the EDA industry has been missing the fundamental components of power management. These include a system-level perspective, modeling of system electronics, and multi-voltage power management of the ICs. Going forward we need tools that take holistic view of the entire chip design. The newer tools should capture the complex, yet fundamental, link between the functional, electrical, thermal, and mechanical aspects of power management.
Sreen Raghavan, President & CEO, Vativ Technologies
TV markets are changing rapidly. The transition happening from analog to the digital TVs is evident worldwide and more so in the developing countries. IPTV, mobile TV, HDTV and several other innovations are set to change the television landscape. Given this scenario, the digital TV market faces a number of transitions, and the semiconductor industry has to address these changes on a global basis. High quality video is the main force driving the adoption of the new and emerging digital TVs.
According to a Frost & Sullivan report, the digital TV shipment in 2006 will achieve 68 million and 97.4 million by 2007. The market scale of digital TV semiconductor in 2005 reached US $4.34 billion, and US $11.5 billion is expected to hit by the year of 2010.
With the proliferation of digital TVs, it is interesting to see how companies that provide digital devices that go into the TV will evolve. There are about 5-7 main chips in each digital TV, including digital signal tuner, digital signal demodulator, MPEG decoder, back-end image processor and EMS memory.
There is tremendous interest among consumers for flat panel televisions. This has further been accelerated with falling prices. The price of an LCD TV is down by 30-40 percent compared to the previous year. However the volumes have grown by 100 percent year-over-year. For semiconductor companies the opportunity exists today to enhance the flat-panel TV audio experience immeasurably.
Balaji Bhakta, President and CEO, inSilica
In the coming years we see a disruptive trend happening as mainstream SoCs make way for the new submicron process such as the 65 nanometer process technology. This will bring about a change in the way semiconductor industry has been executing the standard and complex SoCs over the years. The shift from standard SoCs to submicron SoCs will witness differences among the ‘high value’ processors and the standard semiconductor firms as well.
These 65nm advanced process technology introduce many complex designs, which have to be implemented on the highly composite processors. The advanced processes provide improved drive content, are high in performance and low in cost and creates potential for the complex processes to move into other simpler processes.
The standard SoCs face leakage, poly orientation, which is however, secondary in nature. These issues have a lesser impact on the standard processes in achieving their target.
The performance is not hampered and goals are achievable. But, in the 65nm technology, these are primary factors that need attention. For instance at inSilica, we have developed a strong team with equally strong methodology and proprietary IP that solves the challenges these very deep submicron SoCs face. It can turn into a disaster worth millions if the desired specifications of the chips do not match with the cost goals.
Manju Hegde, Co-founder & CEO, Ageia
The parallel processors, which used to be on high-end servers for scientific/industrial applications, are now available in the consumer space and are being driven by consumer applications.
One of the consumer application, gaming, is now driving High Performance Computing (HPC). They are the ingredients of a Perfect Storm: the market is large and growing, gaming is the harbinger of much of digital interactivity and digital entertainment. There is a clear drive for higher performance from the consumers, the application providers and the vendors. Moreover, there is opportunity for architectural innovation. The slate of applications is rich and vast — financial modeling, video encoding and editing, molecular dynamics, imaging/Image reconstruction, climate modeling, CAD/ CAM applications, physical simulation, etc.
The theoretical performance offered by the new processors is ramping up at an astonishing rate. However, there is a significant entry barrier to each application, which consists of the legacy code of which there are literally hundreds of millions of lines written for sequential processors. The model for exploiting this untapped potential will therefore be a vertical one wherein a core, unified set of applications will need to be targeted and the application design, tools and run-time will all have to be accommodated in a diverse environment which harbors parallel and sequential processors, networked applications, sub-optimal code for related applications and impatient consumers. The winners in this environment will be the companies that can navigate through this tricky environment and can capture consumer attention amidst the murkinesss.
Ravi Pappu, Co-founder, ThingMagic
The year 2007 will see privacy and security as important concerns in the RFID (Radio Frequency Identification) industry. EPC Generation two protocol has better and improved security and a marked advancement over previous efforts. As the development of Generation three protocol is underway, RFID vendors and customers consider stringent security as a critical component.
Users want to be certain that RFID tag data, whether it be personal credit card data, passport information, or consumer packaged goods (CPG) pallet data, is able to be quickly and accurately read by trusted RFID readers and networks. At the same time remain shielded from hackers and other unauthorized users. RFID is deployed in consumer applications; thus, satisfaction of general public is optimum so that privacy and security will not be compromised.
Generation two protocol addresses some of these concerns, which enables RFID tags to be rendered inoperable when there is no need to use a tag. For eg. when an item has been sold and a retailer no longer needs to track it. Generation two further protects each tag’s singular identity by disguising the unique EPC number identifier.
The air interface needs to be secured and Generation three protocol security conceivably include encryption of confidential tag data, tag passwords, and tag pseudonyms that change serial numbers each time a tag is read. Such a proactive approach could stay ahead of any legislation with the same goal—providing cost effective, comprehensive RFID security for RFID users and consumers.
Shiv Tasker, CEO, BlueSpec
The Global market will witness a huge input of products that are low-priced and user-friendly by the end of this year. 2007 will prove to be a pivotal year for the semiconductor industry as service providers and subsidiaries of multi-national companies will have to apply tools, methodologies and processes that will not only be cheaper but faster and better. These new tools will allow engineers to move beyond the abstraction layers to handle more complexity. Also allow changes as per the requirements of the industry. This year, the local market will also witness the emergence of projects that are specially designed and have semiconductors optimized in its costs and functions for the Indian consumer. If popular among the Asian markets, it will bring in simple and useful goods for the consumers.
Increasing demand for hardware engineers will come face to face with the crisis of qualified and experienced engineers.