Efficient Machine Learning on HiFi DSPs Using TensorFlow
This presentation will walk through the details of building a keyword-detection model and deploying it on a HiFi DSP using the TensorFlow tool chain.
Overview
Audio, Voice, and Speech Processing Couldn’t Be Easier
Looking for the best-quality audio-, voice-, and/or speech-processing core for your next design? Need all the software already ported for immediate deployment? With the largest selection of audio DSPs, software packages, and an extensive ecosystem of partners, adding audio to your next SoC couldn’t be easier. Make the choice that more than 150 licensees have already made—select Cadence Tensilica HiFi DSPs for all your audio-, voice-, and speech-processing requirements.
Accelerating Machine Learning Applications on HiFi
In addition to traditional DSP workloads, the Tensilica HiFi DSP is increasingly being tasked with running machine learning applications. The machine learning functions in the world of audio include wake word detection, audio scene detection, speech recognition, and noise reduction. In order to reduce the time to market for these neural network (NN) tasks, the HiFi DSP software offering has been expanded to include new optimized libraries.
Machine learning applications must be energy and memory efficient to be deployed broadly in Tensilica HiFi DSP-based systems. The HiFi DSPs' support for TensorFlow Lite for Microcontrollers allows for timely development of these inference-at-the-edge solutions for resource-constrained embedded systems. The Tensilica HiFi DSPs' optimized libraries ensure optimal performance.
Largest Partner Ecosystem with Expansive Range of Audio Software
Cadence has your audio software needs covered with over 150 partners in its comprehensive audio ecosystem. These companies range from industry giants like Dolby and DTS to smaller companies providing innovative sensor fusion, always-on, sound-enhancement, and noise-reduction software.
We have more than 300 audio, voice, speech recognition, and voice enhancement software packages already ported to the HiFi DSP architecture. This means you can get up and running very quickly, and can easily port your own proprietary software, completely in C, while also maintaining or surpassing the performance of assembly on other DSPs.
Key Benefits
Market Leadership
Tensilica HiFi DSPs are widely adopted by OEMs worldwide, with billions of HiFi DSP cores embedded in a vast array of products
Optimal Performance, Energy, and Area
Tensilica HiFi DSPs provide high performance for high-quality, feature-rich products, energy efficiency for always-on use cases, and a compact footprint to achieve application scalability by integrating multiple HiFi DSP cores
Extensible Instruction Set
Based on the Xtensa® architecture, Tensilica HiFi DSPs are often customized and extended by our customers to accelerate their applications
Neural Network
Tensilica HiFi DSPs accelerate neural network inferencing in addition to traditional DSP workloads, and migrating NN workloads to HiFi DSPs is made easy by the support for the industry-standard TFLM framework
Software Libraries
Tensilica HiFi DSPs come with a vast array of optimized, field-proven, easy-to-integrate software libraries for DSP and neural network functions as well as codecs for music playback and voice communication
Ecosystem of 150+ Partners
OEMs can offer product differentiation and accelerate time to market with innovative, ready-to-deploy speech/voice and audio applications from our partners
Target Markets
Products
HiFi DSPs to suit your Audio, Voice and Speech needs
Software
HiFi Nature DSP Library
Optimized math libraries to simplify programming/porting DSP algorithms
HiFi NN Library
Highly optimized functions commonly used in speech NN processing
Cadence Optimized Codecs
Most comprehensive set of audio/speech codecs in the industry
XAF
Software framework to offload and manage audio/voice/NN processing to DSP
TensorFlow Lite Micro
ML Framework
Partners
Cadence, through its Tensilica processor IP, brings together best-in-class products and services from industry leaders to help you accelerate the development of your SoC designs while meeting your demanding power and performance requirements. Browse the list of Tensilica processor IP service partners below.
Technologies
Functional Safety for Automotive
Cadence is committed to enabling Functional Safety applications across the Tensilica processor lineup, whether it's an off-the-shelf processor/DSP IP or a custom domain-specific processor for a specific application
Learn More
Tensilica Processor Technology
Differentiate, reduce time to market, add flexibility, and get the best performance, power, and area
Learn More
TIE
Customize your DSPs/processors for optimized performance, energy efficiency, and differentiation
Learn More
Development Toolchain
Create a perfect match for your application, guaranteed correct by construction
Learn MoreTraining & Support
Need Help?
Cadence is committed to keeping design teams highly productive with a range of support offerings and processes designed to keep users focused on reducing time to market and achieving silicon success.
Free Software Evaluation
Try our SDK Software Development Toolkit for 15 days absolutely free. We want to show you how easy it is to use our Eclipse-based IDE.
Apply Now
Training
The Training Learning Maps link opens in new tab help you get a comprehensive visual overview of learning opportunities.
Training News - Subscribe link opens in new tab
Online Support
The Cadence Online Support (COS) system fields our entire library of accessible materials for self-study and step-by-step instruction.
Request Support
Xtensa Processor Generator (XPG)
The Xtensa Processor Generator (XPG) is the heart of our technology - the patented cloud-based system that creates your correct-by-construction processor and all associated software, models, etc. (Login Required)
Launch XPG
Technical Forums
Find community on the technical forums to discuss and elaborate on your design ideas.
Find AnswersResources