Cadence launches sixth-generation Tensilica HiFi iQ DSP for voice AI and immersive audio


SOURCE: NEWELECTRONICS.CO.UK
JAN 22, 2026

News

The sixth-generation Tensilica HiFi iQ DSP Credit: Cadence

The new DSP is based on an updated architecture aimed at meeting the rising computational and energy-efficiency demands of system-on-chip (SoC) designs used in home entertainment systems, automotive infotainment platforms and smartphones. These markets are increasingly relying on on-device AI for voice interaction, immersive audio and real-time signal processing.

According to Cadence, the Tensilica HiFi iQ DSP delivers twice the raw compute performance of its HiFi 5s predecessor, alongside up to eight times higher AI processing capability. The company also reports energy savings of more than 25 per cent across most workloads, and performance gains exceeding 40 per cent for a range of audio codecs.

“With the increased growth of AI across virtually every facet of the industry, audio and language have become critical interfaces for users,” said Anshel Sag, 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 position it to enable future AI?enhanced applications and offload processing from other IPs.”

The demand for higher-performance audio processing is being driven by several trends, including immersive and object-based audio formats, higher sampling rates, advanced natural language processing, and features such as automotive road noise cancellation. These use cases require expanding AI and machine learning capabilities, while maintaining low latency and low power consumption—particularly in edge AI applications where power budgets are tightly constrained.

Cadence said increasing clock frequencies or adding multiple cores is often impractical in these environments, prompting the development of a new architecture focused on AI?enhanced audio workloads. Compared with the HiFi 5s DSP, the HiFi iQ introduces integrated support for formats such as FP8 and BF16, improved auto-vectorisation to simplify programming, and broader support for modern voice AI models.

The company said these changes enable more efficient execution of immersive audio codecs including Dolby MS12, Eclipsa Audio, Opus HD and Audio Vivid. Signal processing tasks such as keyword spotting, active noise cancellation, beamforming and automatic speech recognition can also be handled concurrently, enabling more advanced natural language processing on-device. Support for multi-stream and multichannel audio further allows for 3D spatial rendering and zoned audio experiences.

Boyd Phelps, senior vice-president and general manager of Cadence’s Silicon Solutions Group, said the DSP reflects a broader shift towards voice-first interfaces. “With recent advancements in LLMs, SLMs, energy-efficient SoCs and on-device AI, voice input is poised to become the new keyboard,” he said. “To enable seamless interaction with the latest physical AI applications, SoC providers need energy-efficient and easily programmable IP capable of running SLMs on-device while performing voice and audio processing seamlessly.”

He added that Cadence’s long-standing presence in the audio and DSP market positions it to continue improving performance and energy efficiency as AI workloads expand.

The Tensilica HiFi iQ DSP is designed to run both small and large language models directly on the DSP, allowing it to function as a standalone processor for voice AI tasks. For additional performance, it can be paired with Cadence’s Neo NPUs or custom neural processing units developed by customers.

The DSP is supported by the Cadence NeuroWeave software development kit, as well as frameworks including TensorFlow Lite for Micro, LiteRT and ExecuTorch. Cadence said developers can also draw on an established ecosystem of software libraries, compilers and codec packages to speed deployment.

The Tensilica HiFi iQ DSP is expected to be available to lead customers and partners in the first quarter of 2026, with general availability planned for the following quarter.

Artificial Intelligence

Audio Technology

Electronics

ASIC & SoC

http://www.cadence.com