Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > Memory/Storage

UI as the next battlefield in CE (Part 2)

Posted: 05 Apr 2012 ?? ?Print Version ?Bookmark and Share

Keywords:predictive intelligence? cloud computing? general-purpose processor?

The challenge for infotainment OEMs is to balance the need to support "in car" applications that primarily use the embedded resources within each car's equipment, with drivers' need to interact with the outside world. On one hand, there are multiple functions that will likely remain critical to the operation and safety of the vehicle and its driver, such as volume control, multimedia features or mobile phone access. On the other hand, consumer demand remains high for access to maps or social networking resources. As such, many OEMs are employing a hybrid approach utilizing onboard processing dedicated to resources for key in-car functionality and cloud computing on an "as needed" basis for less-frequently used and advanced functions. This approach allows the OEMs to introduce systems that are both high performance and reasonably priced.

The UI processor
Embedded systems are increasingly demanding high performance processing capabilities. This is particularly visible in the automotive infotainment segment, where competition among OEMs is driving innovation and the adoption of advanced computing functionality to support multicore architectures and 3-D graphics capabilities. At the same time, multiple other functions within the automotive environment continue to compete for available computing resources. This burden is centered on the applications processor, which is the main computing engine and is usually heavily burdened by numerous functions such as 3D graphics and advanced communications protocols. As a result, these embedded systems have limited resources available to dedicate to other functions.

A good example of the resource limiting nature of infotainment systems and its impact on key functionality is the user interface. Advanced speech recognition, for example, requires both high MIPS and fast memory access. While system designers strive to produce a highly optimized infotainment system that supports multiple advanced functions combined with an elegant UI, they remain constrained by the available technology, specifically hardware that could support their design requirements. UI increasingly is requiring access to advanced look-up tables and algorithms, which end up in direct competition with other system functions. Finally, the requirement of executing multiple recognition algorithms on a general-purpose processor remains a hindrance to the low power requirements of these systems. Even though advanced applications processors are rapidly shifting from single- to dual- and quad-core technologies for high-end infotainment systems, we believe the UI will remain a compute intensive function, which will require separate hardware accelerators and flexible software algorithms.

Figure 1: The UI processorAdvanced system intelligence requires sophisticated analysis capabilities involving hardware accelerators, flexible software algorithms, and look-up tables. Single-feature ICs will quickly evolve into dedicated UI processors to implement and offload processing of multiple forms of recognition from the applications processor.

As infotainment OEMs continue to integrate advanced human machine interface (HMI) capabilities, such as touch and voice recognition, the UI function will remain a key system bottleneck. Among the UI bottlenecks and design challenges OEMs must address are the ability to process sophisticated software algorithms, accurately operate in various high noise environments like the confined cabin of a car, support privacy requirements for data security and access to premium/paid content, protect devices from malicious hacking, and support real-time off-board interaction with location-aware services and advertising. As additional forms of UI are introduced in these systems, next generation systems will face the added challenge of implementing multiple forms of recognition C speech, voice, image, facial, and emotion C within the same system (figure 1). As the real-time requirements of a single UI technology continue to disrupt the reliable operation of an applications processor, the introduction of additional UIs will only serve to exacerbate this issue, as these added functions compete for shared resources.

?First Page?Previous Page 1???2???3?Next Page?Last Page

Article Comments - UI as the next battlefield in CE (Pa...
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top