Voice-user interfaces will be driven by programmable silicon

May 20, 2016 // By EDN Europe
Huw Geddes, Director of Marketing, XMOS
If you believe in Makimoto’s Wave, which charts the swing between fixed function chips and programmable solutions over the decades, we’re about to move into a cycle where programmability dominates.

It also coincides with a key stage of the evolution of the Internet of Things where many different categories of electronic device will be conceived and developed, instead of one dominant product category like PCs or cell phones. Many of these devices will use voice interfaces with always-on connections to networks of local servers or Cloud services, instead of (or in addition to) the graphical interfaces we’ve become accustomed to.

To achieve these new product categories, developers need solutions that provide high performance control, DSP and outstanding IO flexibility. Time to market and opportunities to differentiate products will be critical as designers fight to gain attention. The ability to innovate in the processing algorithms and the implementation, both in the client and in the cloud, will be essential. New standards will have to be supported, and products will have to be upgradeable so they don’t quickly become obsolete.

xCORE multicore microcontrollers, developed by XMOS, are uniquely applicable for these diverse voice enabled IoT applications, providing single-chip solutions to a wide range of control and interfacing applications. The xCORE processor contains a series of logical cores that provide high performance, control and DSP processing with fast deterministic IO response for implementing flexible IO structures. This provides the advantages of a fixed function implementation with dedicated signal chains and the flexibility of a programmable engine. Large, tightly coupled memories and optimised DSP libraries provide developers with excellent performance for algorithms requiring intensive random memory accesses such as fast FIR algorithms for analysing data from Pulse Density Modulation (PDM) microphones. Algorithms such as fast FIR show similar or even better performance on xCORE architecture compared to dedicated stand-alone DSP devices.

Programmability brings other advantages for the designer. Signals can be processed locally to save power and faster response times. Rather than sending raw data over the broadband link, the data can be analysed locally for a faster, more responsive implementation. Further semantic processing in the cloud can be used to improve the quality of the requests, storing and aggregating large amounts of data and parameters that can be downloaded to modify the local software.    

One challenge with programmability is providing the level of performance at the same time as delivering the flexibility at the right cost point. Once that has been achieved, innovation moves even faster with new algorithms leading to improved performance and reduced power. For applications based on natural speech interaction such as connected assistants, the sub-millisecond latency enabled by the xCORE architecture provides that performance.

The pendulum associated with Makimoto’s Wave will swing back to fixed function in due course, but for now IoT applications with voice-user interfaces need programmable solutions to help them evolve. xCORE is ready to be part of this next generation of smart products.

 

Huw Geddes

Director of Marketing, XMOS

Huw Geddes has an extensive background in the delivery of technology to designers, developers and engineers. Prior to joining XMOS as a Information/Documentation Manager, Huw worked as Technology Transfer Manager at the 3D graphics company Superscape Ltd, and Technical Author at VideoLogic Ltd.