【Application Solution】 Ultra-Miniature Packaging + AI Noise Reduction + Ultra-Low Power Consumption – AWINIC One-Stop Audio Solution Unlocks New Audio Experiences for Wearable Devices

2026-07-16

PART 01
I. Industry Market Pain Points: Core Audio Bottlenecks in AI Wearable Devices

AI glasses, action cameras, and lightweight smart wearable devices are rapidly gaining popularity, but the audio chain commonly suffers from design bottlenecks that are difficult to overcome:

1. Extremely Constrained Structural Space
The PCB area inside AI glasses temples, action camera bodies, and portable wearable devices is at a premium. Traditional power amplifiers and voice Codec chips come in relatively large packages with numerous peripheral components, crowding out space for battery and main controller layout.

2. Prominent Battery Life Pressure
Wearable devices typically have batteries of ≤300 mAh. Traditional audio chips have relatively high quiescent power and playback power consumption, while voice wake-up and real-time loudspeaker output significantly reduce overall device battery life.

3. Poor Loudspeaker Audio Quality
The speaker cavities in wearable devices are extremely small, resulting in missing low frequencies, insufficient volume, airflow noise, and severe distortion and clipping. Conventional power amplifiers lack built-in audio algorithms and rely on additional main controller computing power for parameter tuning.

4. Outdoor Voice Interaction Failure
In scenarios involving sports, cycling, strong winds, and crowded marketplaces, wind noise and environmental noise interference are extremely strong. Traditional voice chips have high VAD wake-up power consumption and weak noise reduction capabilities, causing offline voice recognition rates to drop sharply.

5. High Software and Hardware Development Costs
Power amplifiers and voice processing chips belong to two separate solution sets, algorithms require third-party licensing, debugging cycles are long, main controller computing power is heavily occupied, and the implementation barriers are high for small and medium-sized manufacturers.

AWINIC's Digital Series power amplifiers + AI-NPU voice chip combo solution addresses all pain points from three layers – hardware packaging, underlying power consumption, and self-developed DSP/NPU algorithms – making it suitable for mass production of lightweight wearable devices.

Figure 1 – AWINIC Full-Chain Solution

PART 02
II. AWINIC Integrated One-Stop Solution Architecture

AWINIC provides an integrated one-stop audio solution with no additional third-party algorithm chips, a compact PCB layout, and suitability for miniaturized wearable designs.
Microphone Array → Uplink NPU (VAD Wake-up + NPU Noise Reduction + Voice Recognition) → I2S Digital Audio Bus → Digital Series DSP Audio Amplifier (Audio Enhancement + Loudspeaker Driving) → Miniature Speaker

Figure 2 – AWINIC Integrated Audio Solution

Figure 3 – Three Core Advantages

PART 03
III. Core Chip Product Specifications

(I) AW88103CSR Ultra-Miniature Flagship Power Amplifier

1. Miniaturized Packaging
WLCSP wafer-level package with dimensions of only 1.77 mm × 2.175 mm – the industry's smallest package for the same power rating. It reduces placement costs and PCB routing complexity and can fit into the narrow space of AI glasses temples.

2. Hardware Specifications
6.25 V adaptive boost, 2.1 W output into an 8 Ω load, supports receiver/loudspeaker dual modes, I²C control interface; output noise floor as low as 10 μV, with strong RF immunity that effectively eliminates TDD RF noise in wearable devices.

3. DSP Algorithm Capabilities
Integrated with the Feitian DSP™ core, featuring AGC automatic gain control, anti-clipping limiting, speaker protection, and airflow noise suppression – a full suite of algorithms. No need for the main controller to handle audio effect computations; the small-cavity speaker automatically has its low frequencies enhanced, delivering clear, full-bodied vocal reproduction.

4. Ultra-Low Power Design
Adaptive boost with dynamic voltage regulation reduces playback power consumption by 25% at low volume levels; built-in multi-level power-saving modes with microampere-level quiescent current in standby, suitable for all-day wearable scenarios.

5. Protection Mechanisms
Fourfold hardware protection – short circuit, over-temperature, undervoltage, and overvoltage – suitable for outdoor action cameras in harsh, bumpy environments with wide temperature variations.

6. Stable Operation Across Temperature Ranges
Operating temperature range of –40°C to 85°C, suitable for outdoor action cameras and all-weather wearable devices.

(II) AWA89601 & AWA89501 – NPU Uplink Voice Processing Chips

Integrates AI offline voice interaction capabilities, supports edge-side voice interaction in multiple languages, incorporates AI deep learning noise reduction and echo cancellation functions, and supports high-definition voice calls for applications in laptops, smart homes, security systems, AI toys, and other devices.

Figure 4 – ANPU Application Block Diagram

1. Five Core Voice Processing Technologies
Five core voice processing technologies greatly improve the accuracy and naturalness of voice recognition.

Figure 5 – Five Core Voice Processing Technologies

Figure 6 – Industry-Leading Complete Voice Algorithms

2. AI Voice Chip AWA89501QNR – Core Application Scenarios for AI Edge-Side Voice Interaction

Scenario 1: AI toys can perform direct voice interaction without requiring a complex internet connection process.
Scenario 2: Smart home appliances can execute control commands in environments without network coverage.

Figure 7 – AWA89501QNR Solution Advantage Comparison

Applicable Scenarios: AI toys/robots, AIoT devices, smart speakers, industrial control equipment, smart homes.

Figure 8 – AWA89501QNR Applicable Scenario Examples

3. AI Voice Chip AWA89601QNR – Deep Learning Noise Reduction
Integrates AI deep learning noise reduction and echo cancellation functions, supporting high-definition voice calls.

Applicable Scenarios: Walkie-talkies, building video intercoms, AI robots, smart helmets.

Figure 9 – AWA89601QNR Applicable Scenario Examples

AI wearable devices are shifting from being "display hardware" to "voice interaction–centric" products. The miniaturization, low power consumption, high audio quality, and high recognition accuracy of the audio chain have become core selling points for product differentiation.

AWINIC's integrated audio solution, with its four core capabilities – domestically developed self-designed chips, miniature package architecture, built-in software and hardware algorithms, and hardware-level ultra-low-power VAD – addresses all pain points in the full audio chain for wearable terminals such as AI glasses and action cameras in a one-stop manner. It empowers the industry to create differentiated wearable products featuring lightweight design, long battery life, powerful voice interaction, and superior loudspeaker audio quality.