By Dr. John Heinlein
There’s no denying the pervasiveness of mobile devices and mobile communications technologies. These applications are obviously changing how we communicate and share information, but also increasingly how we do business. And consumers keep expecting more from their mobile devices. Today, the performance and capability in high-end smartphones and certainly tablet computers are getting on a par with notebook computers. Consumers who previously didn’t know or care about technical specifications are being barraged with milliwatts, gigahertz, and nanometers. Form factors of devices are becoming more and more critical to differentiating a device, where new devices are measured in millimeters and grams compared to prior generations. And of course, there is the pervasive need for more battery life. The irony is that making devices smaller and thinner means less volume (and mass) for the lithium ion-based chemical batteries that are today’s standard energy carrier. Consumers still expect their devices to do more and also last longer when running, as well as when they are on stand-by.
These increasing expectations for mobile present opportunities as well as technical challenges that we in the semiconductor industry must address.
My wife will attest that I have an insatiable sweet tooth, and so it’s not surprising that I equate this problem with a layer cake, featuring many layers (chocolate, thank you very much). Like a well-formed layer cake, the problem of delivering leading, next-generation mobile devices is not solved through any one component or technology. It takes a solution with all the layers of the stack working together.
At the highest level, the ARM Connected Community features over 970 partners who deliver value-added capabilities to the ARM partnership. These providers span applications, middleware, tools, IP blocks, design services, and a whole host of other capabilities, and the breadth of the ecosystem supports the richness of capabilities that enables the mobile devices we love.
ARM itself focuses our efforts on a range of technologies that are important to the middle layers of the cake. ARM delivers our industry-leading, energy-efficient Cortex™ processors and Mali™ graphics processors, and we continue to lead the way with continued innovation. We recently rolled out our unique big. LITTLE technology, which gives the best of both worlds of high performance and energy efficiency. We provide CoreLink™ system IP that connects the system together and increases performance and capabilities for sharing information across devices. And finally we deliver Artisan® physical IP that helps designers implement these designs for manufacturing.
The lowest level of the cake involves realizing these conceptual designs into actual silicon chips that power the broad array of mobile devices. This job falls to semiconductor foundries like GLOBALFOUNDRIES, which has partnered with ARM for many years on mobile innovation. The foundry ecosystem provides a proven manufacturing solution for fabless companies, enabling the amazing growth and diversity of devices in the marketplace today. Through this solution, the industry is delivering high-volume 28nm solutions in the market place and driving rapidly towards tapeouts of 20nm devices next year. GF, for its part, is investing heavily in its 20nm-LPM process technology, which delivers a fully scaled metallization compared to the 28nm node. While often the focus in the media is on transistor performance, scaling the metal is critically important as well, as it drives area improvements, cost reduction and power reduction.
With GF’s recent announcement of its 14nm-XM technology, the company is going beyond the 20nm node and looking to the future, and I for one think its approach is pragmatic. It’s critical that the industry continue the pace of innovation and deliver the next wave of transistor capabilities for production in the 2014 time frame. The growing consensus is that devices beyond 20nm will be based on a FinFET transistor, a 3D device where the current flow can be more effectively controlled using a gate on both sides of the conducting channel.
GF has invested in R&D on FinFET devices for years now and it intends to roll out its FinFET-based 14nm-XM (for “eXtreme Mobility”) technology in 2013, targeting production volume in 2014. This process technology is projected to provide significant scaling and improvement in energy-efficiency compared to previous nodes, and has the potential to extend the kinds of scaling that we have enjoyed from Moore’s Law for so many years. Given the complexity of rolling out a new device technology, the approach leverages the metal solution from 20nm-LPM. This reduces one of the major risk factors in rolling out a new technology, since metal scaling is also a key challenge for manufacturability.
GF has also committed to collaborating with ARM to jointly optimize this process technology specifically for next-generation ARM processors and graphics for the mobile market. This joint optimization is delivered in part through the ARM POP™ IP product line for chip optimization and risk reduction, is a key element in assuring customers get to market quickly with new designs. By delivering POP IP for GF’s 20nm-LPM and 14nm-XM technologies, we can assure the pace of innovation continues into the future.
We are excited that the foundry industry is responding with aggressive technologies that will help drive the next generation of manufacturing and look forward to the exciting new products that this technology will enable.