Femtosense introduces AI-ADAM-100: A Compact SiP Empowering Consumer Tech

Femtosense and ABOV Semiconductor have just launched their latest collaboration, the AI-ADAM-100. This new artificial intelligence microcontroller unit (AI MCU) is designed to bring on-device AI features, like voice-based control, to home appliances and other products. But what sets the AI-ADAM-100 apart is its use of sparse AI technology, which allows for immediate and low-power user responses, along with enhanced security and operational stability.

The AI-ADAM-100 combines Femtosense’s Sparse Processing Unit 001 (SPU-001), a neural processing unit (NPU), and ABOV Semiconductor’s MCU to provide on-device deep learning-powered AI voice processing. This means that appliances can now have voice interfaces that understand natural language and allow users to express their intent in various ways. So whether you say “Turn the lights off,” “Turn off the lights,” or simply “Lights off,” the AI-ADAM-100 will understand your command.

But it doesn’t stop there. The AI-ADAM-100 also includes voice/audio cleanup capabilities that process data before it’s sent to the cloud. This not only improves reliability and accuracy but also reduces the amount of data being sent, resulting in lower backend infrastructure costs. It’s a win-win situation.

Sam Fok, CEO of Femtosense, explains the significance of their sparsity-enabling technology: “With sparsity integrated throughout the AI development stack, the AI-ADAM-100 is the first device on the market to fully unlock the advantages of sparse AI.” This technology allows manufacturers to deliver compact and efficient AI processing to a wide range of markets and products, including home appliances, hearing aids, industrial headsets, and consumer earbuds.

But what exactly is sparse AI? Well, it’s a game-changer when it comes to reducing the cost of AI inferencing. By zeroing out irrelevant portions of an algorithm and only allocating resources to the relevant parts, sparse AI can deliver up to a 10x improvement in speed, efficiency, and memory footprint. This means that manufacturers can implement deep learning-based AI models that are up to 100x more powerful without sacrificing performance.

One of the main advantages of the AI-ADAM-100 is its ability to bring AI to edge devices that don’t have the price or power flexibility for a GPU or cloud connectivity. This has been a barrier for many edge applications, but with the AI-ADAM-100, manufacturers can now implement voice language interfaces even on devices that aren’t connected to the cloud.

Furthermore, the AI-ADAM-100 significantly reduces power consumption and backend cloud loading by resolving tasks on-device. For example, home appliance manufacturers can implement wake-up and control functionality that allows other system controllers to consume less power when not in use. This means that the system can listen for voice commands and either process them on-device or wake the system to send the command to the cloud. It’s all about efficiency and power optimization.

The AI-ADAM-100 is the result of a fruitful partnership between Femtosense and ABOV Semiconductor. Choi Won, CEO of ABOV Semiconductor, highlights the strengths of this collaboration: “The AI-ADAM-100 is the best-optimized AI MCU solution for voice and audio-based AI applications and enables a variety of on-device AI applications for consumer electronics and standalone devices.” With their combined expertise, Femtosense and ABOV Semiconductor are committed to developing cost- and power-efficient AI MCUs for global customers.

The AI-ADAM-100 has already been put through rigorous testing under multiple noise conditions and has proven its top-notch voice command recognition performance. This makes it an ideal choice for global home appliance makers who are looking to streamline user experiences by reducing the number of buttons on their devices.

If you’re excited about the AI-ADAM-100, you’ll be pleased to know that engineering samples are already available, with commercial mass production expected later this year. And to support developers, Femtosense provides a range of software tools, evaluation boards, and demo AI models, including Smart Home Appliance Wake-up and Control.

So get ready for a new era of on-device AI with the AI-ADAM-100. It’s time to bring voice-based control to your home appliances and experience the power of sparse AI.

Technology Explained

---

GPU: GPU stands for Graphics Processing Unit and is a specialized type of processor designed to handle graphics-intensive tasks. It is used in the computer industry to render images, videos, and 3D graphics. GPUs are used in gaming consoles, PCs, and mobile devices to provide a smooth and immersive gaming experience. They are also used in the medical field to create 3D models of organs and tissues, and in the automotive industry to create virtual prototypes of cars. GPUs are also used in the field of artificial intelligence to process large amounts of data and create complex models. GPUs are becoming increasingly important in the computer industry as they are able to process large amounts of data quickly and efficiently.

---

NPU: NPU, or Neural Processing Unit, is a type of specialized processor that is designed to handle complex artificial intelligence tasks. It is inspired by the structure and function of the human brain, with the ability to process and analyze large amounts of data simultaneously. In the computer industry, NPU technology is being used in various applications such as speech recognition, image and video processing, and natural language processing. This allows computers to perform tasks that were previously only possible for humans, making them more efficient and intelligent. NPU technology is also being integrated into smartphones, self-driving cars, and other devices, making them smarter and more responsive to user needs. With the increasing demand for AI-driven technology, the use of NPU is expected to grow and revolutionize the way we interact with computers in the future.