Feature: Semiconductors
to dominate demand are shrinking, and there are far more new market opportunities to target – all of which demand unique solutions and, for many manufacturers, making it an unsavvy business model to invest heavily in a single market. Ideally, companies would have
unlimited resources to perfectly tailor hardware to the specific needs of each application. However, given the vast array of uses, achieving this level of customisation is unrealistic. Manufacturers must abide by the budgetary and logistical limitations they face, compromising where appropriate to maximise their resources. In this context, DSPs have become
crucial. Engineers often face uncertainties regarding components and budgets at the beginning of a project. The DSP stands out as a reliable and proven tool for incorporating essential functionalities into designs.
The role of AI When considering DSP solutions, there are two approaches: a dedicated DSP focused solely on signal processing, and a more versatile solution that combines a DSP with other functionalities. The dedicated type will offer higher efficiency and precision in signal processing, but it comes at the price of being harder to integrate, taking up more space within a system and greater cost. This can lead to slower market entry, a higher- priced bill of materials and more complicated design challenges. General-purpose hardware, on the
other hand, offers balance between cost and performance. It allows for flexible integration, and engineers can merge it with other functionalities. The most efficient systems allow for software-based reconfiguration, where modifications do not require physical redesign of the component or the device, thus keeping down costs and development time. Using AI takes this versatility a
An AI-enhanced DSP presents a dynamic, adaptable and potent method of data capture, processing and initiating actions based on those insights
step further. It can augment the DSP capabilities, enabling devices to process and improve signal quality more effectively, even when dealing with noisier inputs. AI’s contribution to DSP extends to more practical applications, such as enhancing voice recognition in devices during video conferencing, for example. By applying machine-learning algorithms, DSPs can adapt a system to its environmental noises, such as wind or background chatter, offering clearer communication. AI can extract insights from data
translated by DSPs, enabling other components around it to act more efficiently on this information. The cloud is a conventional means of doing so, but some disruptive systems will leverage AI to perform on-device data processing, eliminating the need for cloud dependency and thereby enhancing privacy and performance. Of course, not every device that
uses DSP is AI-ready. It’s important to ensure that during the design phase, developers create a system-level design that accommodates the right I/O interfaces, and other parts needed to use AI. There’s no point in committing to an AI-enabled design if it can’t be integrated without compromising the device itself. When a balance is found, the
integration of AI and DSP signifies a
leap forward in how devices interpret and respond to the world around them, paving the way for more intelligent and responsive technologies.
Looking ahead An AI-enhanced DSP presents a dynamic, adaptable and potent method of data capture, processing and initiating actions based on those insights. In future consumer electronics, more means of interacting with a device will be added. For example, everyday systems like
computers, televisions, wearables and others will efficiently recognise voice commands, with AI algorithms delivering much improved accuracy by distinguishing spoken commands from ambient noise. Moreover, AI will introduce more sophisticated capabilities to, say, identify individual speakers and link them with their subscriptions or accounts. AI-driven DSP can thus become an important part of streamlining user interactions and delivering personalised experiences. The automotive sector is set for
similar advancements, with refined voice command technologies enhancing the in-vehicle experience. DSPs will play a crucial role in ensuring quality control within the car’s internal systems. Noise suppression is one example, with AI reducing exterior noise interference, such as wind, road or traffic sounds. In manufacturing, DSP will also
‘ignore’ industrial noises, to enhance the effectiveness of voice commands in factories, contributing to safer and more efficient operations. And, DSPs can contribute to the analytics monitoring industrial machinery, enabling the early detection of the need for maintenance if certain sounds or patterns that indicate a defect are present. It can also combine with I/O controls to trigger emergency stops of a machine in response to specific commands or noises. So, the future of DSPs is not only
bright, but exciting and very promising, too.
www.electronicsworld.co.uk June 2024 21
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52
