Cookie
This website requires cookies to provide all of its features. For more information on what data is contained in the cookies, please see our Privacy Policy page. To accept cookies from this site, please click the Accept button below.
| 供应商 | CEVA |
|---|---|
| 型号 | Heterogeneous and Secure AI/ML Processor Architecture for Smart Edge Devices |
| 类型 | Soft IP |
| 到期日 | In Production |
| 源网址 | View more |
| 特征 | Support wide range of activations & weights data types, from 32-bit Floating Point down to 2-bit Binary Neural Networks (BNN) Unique mixed precision neural engine MAC array micro architecture to support data type diversity with minimal power consumption Out-of-the-box, untrained Winograd transform engine that allows to replace traditional convolution methods while using 4-bit, 8-bit, 12-bit or 16-bit weights and activations, increasing efficiency in a factor of 2x with <0.5% precision degradation Unstructured Sparsity engine to avoid operations with zero-value weights or activations of every layer along the inference process. With up to 4x in performance, sparsity will also reduce memory bandwidth and power consumption Simultaneous processing of the Vector Processing Unit (VPU), a fully programmable processor for handling any future new neural network architectures to come Lossless Real-time Weight and Data compression/decompression, for reduced external memory bandwidth Scalability by applying different memory configuration per use-case and inherent single core with 1-8 multiengine architecture system for diverse processing performance Secure boot and neural network weights/data against identity theft 2 degrees of freedom parallel processing Memory hierarchy architecture to minimize power consumption attributed to data transfers to and from an external SDRAM as well as optimize overall bandwidth consumption |
| 详情页 | View more |
| 附件链接 | View more |
NeuPro-M? redefines high performance AI (Artificial Intelligence) and ML (Machine Learning) processing for smart edge devices and edge compute with heterogeneous and secure architecture.
NeuPro-M is a highly scalable high-performance AI processor architecture with an exceptional power efficiency of up to 24 Tera Ops Per Second per Watt (TOPS/Watt).
NeuPro-M provides a major leap in performance mainly because of its heterogeneous coprocessors that demonstrate compound parallel processing, once within each internal engine and secondly between the engines themselves.
Ranging from 8 Tera Ops Per Second (TOPS) up to 160 TOPS per core and is fully scalable to reach above 1200 TOPS using multi-core instantiations, NeuPro-M can cover a wide scope of edge AI compute application needs which enables it to fit a broad range of end markets including automotive, surveillance, mobile, consumer, industrial & IoT and robotics.
With various orthogonal bandwidth reduction mechanisms, decentralized architecture of the management controllers and memory resources, CEVA's NeuPro-M can ensure full utilization of all its coprocessors while maintaining stable and concurrent data tunneling that eliminate issues of bandwidth limited performance, data congestion or processing unit starvation. These also reduce the dependency in the external memory of the SoC its embedded into.
NeuPro-M builds on CEVA’s industry-leading position and experience in deep neural networks applications. Dozens of customers are already deploying CEVA’s vision & AI platforms along with the full CDNN SDK toolchain in consumer, surveillance and ADAS products.
NeuPro-M was designed to meet the most stringent safety and quality compliance standards like automotive ISO 26262 ASIL-B functional safety standard and A-Spice quality assurance standards and comes complete with a full comprehensive CDNN software stack including:
NeuPro-M system architecture planner tool – Allowing fast and accurate neural network development over NeuPro-M and ensure final product performance
Neural network training optimizer tool allows even further performance boost & bandwidth reduction still in the neural network domain in order to fully utilize every NeuPro-M optimized hardware feature
CDNN compiler & runtime, compose the most efficient flow scheme within the processor to ensure maximum utilization in minimum bandwidth per use-case
The NeuPro-M architecture supports secure access in the form of root of trust, authentication against IP / identity theft, secure boot and end to end data privacy.
NeuPro-M is a highly scalable high-performance AI processor architecture with an exceptional power efficiency of up to 24 Tera Ops Per Second per Watt (TOPS/Watt).
NeuPro-M provides a major leap in performance mainly because of its heterogeneous coprocessors that demonstrate compound parallel processing, once within each internal engine and secondly between the engines themselves.
Ranging from 8 Tera Ops Per Second (TOPS) up to 160 TOPS per core and is fully scalable to reach above 1200 TOPS using multi-core instantiations, NeuPro-M can cover a wide scope of edge AI compute application needs which enables it to fit a broad range of end markets including automotive, surveillance, mobile, consumer, industrial & IoT and robotics.
With various orthogonal bandwidth reduction mechanisms, decentralized architecture of the management controllers and memory resources, CEVA's NeuPro-M can ensure full utilization of all its coprocessors while maintaining stable and concurrent data tunneling that eliminate issues of bandwidth limited performance, data congestion or processing unit starvation. These also reduce the dependency in the external memory of the SoC its embedded into.
NeuPro-M builds on CEVA’s industry-leading position and experience in deep neural networks applications. Dozens of customers are already deploying CEVA’s vision & AI platforms along with the full CDNN SDK toolchain in consumer, surveillance and ADAS products.
NeuPro-M was designed to meet the most stringent safety and quality compliance standards like automotive ISO 26262 ASIL-B functional safety standard and A-Spice quality assurance standards and comes complete with a full comprehensive CDNN software stack including:
NeuPro-M system architecture planner tool – Allowing fast and accurate neural network development over NeuPro-M and ensure final product performance
Neural network training optimizer tool allows even further performance boost & bandwidth reduction still in the neural network domain in order to fully utilize every NeuPro-M optimized hardware feature
CDNN compiler & runtime, compose the most efficient flow scheme within the processor to ensure maximum utilization in minimum bandwidth per use-case
The NeuPro-M architecture supports secure access in the form of root of trust, authentication against IP / identity theft, secure boot and end to end data privacy.
FAQ - Frequently Asked Questions
Welcome to Sparsh\'s FAQ! Here you\'ll find answers to our most asked questions. Still have something we should know? Send us an email, we\'re happy to help!
Q1: How can I change my shipping address?
A1: Consectetur morbi, suscipit donec semper vitae sed a class vivamus. Sodales montes porttitor adipiscing nisl sit dui sem fringilla elit. Sagittis lacinia montes nisl mollis lobortis cras nisi. Conubia montes odio taciti magnis morbi mauris lorem pulvinar mollis aliquam. Faucibus facilisi tempus tincidunt eu laoreet. Porta donec vitae suscipit habitasse fermentum vivamus!
- Eget est ad potenti primis id rhoncus vestibulum vestibulum. Ante est vel mattis mattis. Vel eleifend auctor lorem, odio proin quisque potenti parturient euismod. Tristique massa quis morbi netus magna pretium laoreet. Inceptos eget massa ac lacinia vitae suspendisse orci nascetur vel torquent. Feugiat eleifend eget aenean facilisi sapien proin leo dictumst semper orci ipsum? Ultrices duis tellus consequat nostra.
Q2: How do I activate my account?
A2: The instructions to activate your account will be sent to your email once you have submitted the registration form. If you did not receive this email, your email service provider’s mailing software may be blocking it. You can try checking your junk / spam folder or contact us at magento@sparsh-technologies.com
Q3: How do I cancel my orders before I make a payment?
A3: After logging into your account, go to your Shopping Cart. Here, you will be able to make a payment or cancel your order. We will not give refunds if payment is verified.
Q4: How long will it take for my order to arrive after I make payment?
A4: Members who ship their orders within Inida should expect to receive their orders within 2 working days upon payment verification depending on the count of orders received.
If you experience delays in receiving your order, contact us immediately and we will help to confirm the status of your order.
Q5: What are the payment methods available?
A5: At the moment, we only accept Credit/Debit cards and Paypal payments.
Q6: How do I make payments using Paypal? How does it work?
A6: Paypal is the easiest way to make payments online. While checking out your order, you will be redirected to the Paypal website. Be sure to fill in correct details for fast & hassle-free payment processing. After a successful Paypal payment, a payment advice will be automatically generated to Samplestore.com system for your order.
It\'s fast, easy & secure.