Tensilica has announced its HiFi 3 audio/voice DSP IP core for SoC design. The fourth-generation audio DSP is designed to offer higher performance with lower power for high-performance audio post-processing and voice processing algorithms used in smartphone and home entertainment systems.
HiFi 3 features a dual 24- and 32-bit architecture, which enables it to achieve an 80% increase in performance for FFT, FIR, and IIR functions in audio pre- and post processing, plus a greater than 1.5X improvement for most voice codecs compared to the company's HiFi EP. The performance boost is claimed to help significantly reduce the power consumption requirements for leading edge algorithms.
The HiFi 3 is upwards software compatible with the company's HiFi 2 and HiFi EP audio/voice DSPs. In addition, it has been optimized to efficiently handle multichannel audio software packages, such as DTS-HD Master Audio for Blu-ray Disc, which it can support at 233 MHz. For speech it offers direct support for common ITU operations enabling a claimed greater than 50% performance improvement on many voice codecs.
Availability: General availability is expected in March.
Tensilica, Inc., (408) 986-8000, http://www.tensilica.com/.
HiFi 3 Audio DSP information page
HiFi 3 Audio DSP product brief (PDF)