What's next for semiconductor technology?
Current technologies have been stretched nearly to their physical limits, making new solutions costly and challenging. How will the industry meet the increasing global demand for innovative applications? At the panel discussion held during the 2014 imec technology forum USA, ASML, Brewer Science, imec, Lam Research and Nvidia put forward solutions.
From Left to right: Luc Van den Hove, John Chen, Rick Gottscho, Christophe Fouquet and Terry Brewer
While always characterized as a cyclical business of ups and downs, the semiconductor industry has always successfully rebounded with new technologies that have buoyed the next generation of electronic products. Now, however, the industry stands at an inflection point. While global demand for semiconductors continues to rise, the costs associated with fueling the next generation of technology is also growing exponentially. If nothing is done and semiconductor technology continues as it is today, innovation will stall. New markets that can benefit from technology may stagnate. The infrastructure necessary for improving lives"¦ from healthcare to food protection to transportation to entertainment and communications may be stifled.
Achieving the massive changes that took computers outside the office and into pockets, and phones unplugged from walls to be a computer, camera, calendar and game box has derived from research and development that has successfully shrunk the time between new product generations from years down to months.
The ability to deliver new devices that were smaller, denser, faster and cheaper came thanks to three significant drivers: the industry's ability to scale technology to ever finer geometries following "Moore's Law" "“ the roadmap for doubling the number of transistors in a chip approximately every 18-to-24 months; the ability to utilize CMOS as the core manufacturing process platform; and the consumer's insatiable demand for open instant access to communications and information any time and any place.
According to imec president and CEO Luc Van den hove, these formulas for the industry's success have changed. Device structures are pushing the limits of physics, potentially preventing Moore's Law to continue in a reliable progression. Chip design is more complex than ever before, requiring more scrutiny, analysis and testing before manufacturing can even begin. Intricate manufacturing requirements make process control more difficult, making a central neutral platform more important than ever. The cost of bringing new technologies to market is taking longer and costing more than ever before, stretching R&D resources, budgets and time. The appetite for device speed requires materials beyond what silicon can deliver. The cost of manufacturing equipment and the foundries to house it has risen exponentially, shutting out many smaller companies and forcing consolidation and attrition in the numbers of equipment suppliers, foundries, and even in the supply chain. The cost of new designs and incorporating design changes for new generations of chips is rising.
Terry Brewer of Brewer Science agrees that cost and time changes associated with the "Moore's Law" concept are indeed factors challenging the industry. The extraordinary investments required for new solutions is staggering, exacerbating the consolidation occurring in all sectors of the industry. However, he adds that industry and technology challenges are not the only factors in the industry's dilemma. While the industry's maturity and consolidation are factors, he sees changing world economics and social focus outside the industry as a main cause. The factors we see inside the industry are more effects than causes of the inflection.
Richard Gottscho of Lam Research notes that cost has become a general business issue. He states that it is no longer good enough to have the best technology, a company must offer the best technology at the lowest cost.
Gottscho sees IP protection as a key challenge to be overcome if collaboration between "all parts of the food chain" is to enable the industry to move forward through the current inflection challenges.
Taking a more functional approach, Christophe Fouquet of ASML says his company uses a holistic approach to enlarge, calibrate and control the process window. By examining the process "˜fingerprint' at the wafer level, ASML looks to control on-product wafer metrology; calibrate the lithography process and make corrections, all while maintaining process integrity. By integrating metrology, ASML can drive on-product performance for overlay, focus and chemical deposition (CD), while slicing cycle times to minutes. By focusing on doing what they do best, ASML believes it can serve the industry by taking technology to the next level of innovation.
John Chen of Nvidia first described that an inflection point in calculus is defined as where the second derivative of a continuous curve has changed its sign. He then showed that the rate of improvement in IC performance has indeed changed from negative to positive as the technology moves from 20nm planar to 16/14nm FinFET in the industry, making the 20nm a true inflection point. Moreover, the double patterning required to print the 20nm designs has caused a discontinuity in die cost. He showed a simple equation that governs the die cost and chided the industry to increase manufacturing yield as it's the most profound way to reduce die cost and benefits all the parties involved. He stated that other obvious solutions are EUV or other innovative lithography to mitigate the need of multiple patterning thereby getting us back to the die cost curve and/or 450mm wafers, both developments only need a few players to collaborate. Enlarging wafer size to 450mm is in fact a positive discontinuity that is long overdue and by migrating to 450mm, he maintains that the industry can deliver new innovations in addition to increase more dice per wafer. At the end he redefines IC as Innovation plus Collaboration and exhorts the industry to innovate everywhere and collaborate as the metallization in an integrated circuit that interconnects all the individual innovative transistors to make them play in concert. This is in indeed what imec and SEMI bring to the industry.
All agree, however, that scaling "“ the essence of Moore's Law "“ has become a major challenge, forcing R&D costs to increase as the industry works collectively to develop solutions. The issue is simple: traditional methods of solving problems no longer work as they used to, so cost pressures increase, resulting in supplier and customer consolidation. Gottscho noted that, while the vertical integration model (ie, offering design, manufacturing, packaging, assembly and test) has waned, it has been replaced with horizontal integration focused on fabrication process, metrology and other manufacturing issues.
The solution to the challenges at hand is a straightforward one: R&D, the heart of innovation is vital to the industry's successful passage through its inflection point. Through collaborative R&D, the industry can develop technologies for new applications that reach beyond today's traditional semiconductor markets.
Despite these challenges, Van den hove holds great optimism for the industry. The success of mobile devices has only fueled consumption of and demand for semiconductor-based products. Mobile applications will continue to stimulate data consumption from 4G to 5G, as consumers clamor for greater data availability, immediacy, and access. Beyond the traditional computing and communications applications loom new markets for wearable electronics. Nanoelectronics enable disruptive innovations in healthcare to monitor, measure, analyze, predict and prevent illnesses.
Wearable devices have already proven themselves in encouraging healthier lifestyles. The industry's challenge is to ensure that the data delivered via personal devices meets medical quality standards. R&D efforts here will continue to focus on ultralow-power multi-sensor platforms.
Gottscho recommends that the industry take its focus away from the practicalities of scaling and focus instead on solutions by identifying features and benefits that must be included in next-generation systems, and by understanding the requirements for cost, speed, power, form factor and other attributes. He also sees the introduction of new standards as a means to take some cost out of the entire ecosystem. He sees imec as helping to drive innovation and suggests that the industry organization, SEMI, play a larger role in supporting companies. He sees more collaboration between all parts of the food chain needed.
Chen maintains that innovation and collaboration will always lead the semiconductor industry to the next successful generation. Brewer added to this by cautioning that there will always be hindrances as the industry moves forward to a new generation because progress is not attained in a straight line.
Brewer asks the industry to decide if the product diversity offered by "internet of everything" will replace "˜the next big thing' model of innovation.
Luc Van den hove answered the question by highlighting imec's formula for creating R&D teams amongst companies. Imec pools companies from every sector of an industry ecosystem to study all aspects of a topic. In its "˜Internet of healthy things' program, participating R&D partners represent every sector in the healthcare ecosystem -- pharma, diagnostics, medical device manufacturers, foundations, medical researchers, biotech companies and hospitals. In this way imec and its partners can build a new business model for the healthcare industry.
Van den hove adds that other application-specific programs are being shaped in the same holistic way while building on traditional CMOS core process technology to identify new solutions for a myriad of devices using ultralow power.
The challenges are clear:
While there are many facets to the inflection point conundrum, clear answers appear. The high cost of finding solutions and the risk of choosing the wrong solution and the limits of one company's resources will only continue. Through collaboration, companies can develop solutions, explore and create new technologies, shorten times to market, and be ready to bring a new generation of products to a waiting world. The industry may indeed stand at an inflection point, but the future is bright. Innovation cannot be stifled. And collaboration remains the consensus of an industry focused on the next new thing. Today, IC does not just stand for Integrated Circuit; it indeed calls for Innovation and Collaboration.
