SISTC & XMOS Collaborative Development Solution: Innovations in Intelligent Voice Systems with WBC Series MEMS Microphones

——Deep Integration of Hardware and Algorithms, Ushering in a New Era of Voice Interaction

Background and Collaboration Opportunity

With the explosive growth of high-precision voice interaction demands in smart home, automotive voice, and conferencing systems, the co-optimization of MEMS microphones and voice processing algorithms has become a core industry focus. SISTC’s WBC Series MEMS Microphones, featuring high signal-to-noise ratio (70dB+), wide frequency response (20Hz–20kHz), and ultra-low power consumption, have secured significant market share in consumer electronics. To further expand into high-end applications, SISTC has partnered with XMOS, a UK-based leader in audio processing technology, to launch the “WBC-XMOS Intelligent Voice Joint Solution”. This solution integrates hardware design with algorithm processing capabilities, delivering full-stack support for far-field voice recognition, multi-scenario noise reduction, and beyond.

Core Technical Highlights

1. Hardware Layer: Seamless Integration of WBC Microphone Arrays and XMOS Processors

Optimized WBC Microphone Design: Utilizes dual-backplate MEMS technology with PDM output, achieving a 120dBSPL acoustic overload point to capture clear voice signals in complex acoustic environments.
XMOS XVF3800 Voice Processor: The multi-core XVF3800 delivers real-time processing latency as low as 2ms, integrating adaptive beamforming, multi-channel acoustic echo cancellation (AEC), and dynamic noise suppression (DNS) algorithms for precise target sound source localization.
Co-Optimized Interfaces: High-efficiency data exchange between WBC arrays and XVF3800 via I2S/I2C interfaces, supporting linear or circular array layouts for diverse device form factors like smart speakers and conferencing systems.

2. Algorithm Layer: Scenario-Specific Intelligent Processing Engine

Dynamic Beam Tracking: Combines XMOS’s Direction of Arrival (DoA) technology to track speaker positions in real time, maintaining high recognition accuracy even in multi-speaker or mobile scenarios.
Dual-Mode Noise Reduction: Leverages WBC’s high SNR and XMOS’s hybrid noise reduction algorithm (suppressing steady-state and non-stationary noise) to improve voice clarity by over 30% in noisy environments.
Low-Power Wake-Up Mechanism: Synergy between WBC’s low-power mode (<100μA) and XMOS’s sleep mode reduces standby power consumption by 50%, extending battery life.

3. Development Support: Rapid Deployment Toolchain Ecosystem

XMOS VocalFusion Development Kit: Pre-integrated WBC microphone reference designs compatible with Amazon AVS, Google Assistant, and other mainstream voice platforms, shortening product development cycles to 6–8 weeks.
MATLAB/Python Algorithm Validation Toolkit: Open SDKs for beamforming and noise reduction algorithms enable custom parameter tuning for vertical applications like industrial inspection and medical voice recording.

Key Application Scenarios

Smart Home Hubs: A 4–6 mic circular array with XVF3800’s far-field pickup enables accurate voice command response within 5 meters, ideal for smart speakers and home robots.
In-Car Voice Assistants: WBC’s wide temperature stability (-40°C to 85°C) and XMOS’s engine noise suppression enhance voice interaction in driving environments.
Conferencing Systems: Linear arrays paired with XMOS’s full-duplex communication technology eliminate echoes and enable 360° sound localization for video conferencing devices.

Future Vision: AI-Driven Evolution

SISTC and XMOS are exploring next-gen solutions with embedded AI capabilities:

On-Device Voice Models: XMOS’s XCORE-VOICE platform combined with WBC’s high-fidelity input supports localized large language models (LLMs) for real-time translation and semantic understanding.
Thermal and Efficiency Optimization: Leveraging XMOS partner xMEMS’s chip-level cooling technology (e.g., XMC-2400 µCooling) to address thermal challenges in high-compute scenarios.

Conclusion
The SISTC-XMOS collaboration delivers deep integration of hardware performance and algorithmic intelligence while accelerating customer innovation through an open development ecosystem. Moving forward, the partnership will continue advancing MEMS and voice processing technologies to create more natural and reliable voice interaction experiences globally.

For technical details or development support, visit: www.sistc.com/wbc-xmos