“As AI continues to grow in virtually every aspect of industry, audio and language are becoming critical interfaces for users,” said Anshel Sugg, vice president and principal analyst at Moor Insights and Strategy. “The AI performance and energy savings with Cadence’s new flagship Tensilica HiFi iQ DSP enable future AI-enhanced applications and processing offload from other IP.”
Many factors are driving the need for greater compute, including the need for richer experiences enabled by modern immersive audio codecs, audio processing at higher sampling rates, object-based rendering, seamless natural language processing (NLP), and automotive road noise cancellation (RNC). These applications require enhanced AI/ML capabilities and enhanced audio/voice pre- and post-processing capabilities, and on-device processing for edge AI and physical AI applications requires more computation performed with low latency and low power. Most of these designs are power constrained, so increasing the core operating frequency or integrating multiple cores is often not a viable option.
Cadence has leveraged over 20 years of audio solution development and instruction set optimization expertise to develop a new architecture specifically designed for high-performance, AI-enhanced audio workloads. Tensilica HiFi iQ DSP addresses the latest audio market trends and delivers significant improvements over Tensilica HiFi 5s DSP.
- 2x better raw compute performance and 8x better AI performance for advanced compute-intensive audio standards
- Highly energy-efficient on-device AI processing delivers over 25% energy savings across a variety of workloads
- Enhanced auto-vectorization makes programming easier and speeds time to market.
- Integrated support for FP8, BF16, and other formats to run cutting-edge voice AI models
“Recent advances in LLM, SLM, energy-efficient SoCs, and on-device AI are poised to make voice input the new keyboard. To enable seamless interaction with modern physical AI applications, SoC providers need energy-efficient, easily programmable IP that can perform SLM on-device while seamlessly performing voice and audio processing,” said Boyd Phelps, senior vice president and general manager of the Silicon Solutions Group at Cadence. “As the market leader in DSP for audio, voice, speech and AI applications with billions of Tensilica HiFi DSPs in a wide range of embedded products, SoC and system vendors trust Cadence to continue pushing the boundaries of performance and energy efficiency.”
Customer and partner support
Availability and related resources
Tensilica HiFi iQ DSP is expected to be available to key customers and partners in the first quarter of 2026, with general availability in the next quarter. The HiFi iQ DSP is also ISO 26262 Functional Safety (FUSA) certified, making it suitable for safety-critical applications. In the future, HiFi iQ DSP support for cache-coherent multicore configurations is planned.
Cadence is a market leader in AI and digital twins and a pioneer in the application of computational software to accelerate engineering design innovation from silicon to systems. Our design solutions, based on Cadence’s Intelligent System Design™ strategy, are essential for the world’s leading semiconductor and systems companies to build next-generation products, from chips to complete electromechanical systems, serving a wide range of markets, including hyperscale computing, mobile communications, automotive, aerospace, industrial, life sciences, and robotics. In 2024, Cadence was recognized by the Wall Street Journal as one of the world’s 100 best-run companies.
