Basedon the AI-specific APiM framework, a modular deep neural network learning accelerator without any external caching can be used for high-performance edge computing, as a vision-based deep learning computing and AI algorithm acceleration.
Specification:
NPU: LightspeeurSPR2801S (28nm process, unique MPE and APiM architecture)
Peak: 5.6TOPs@100MHz
Low Power: 2.8TOPs@300mW
Platform: ApplicableROC-RK3399-PC platform
Framework : Support Pytorch, Caffe framework, follow-upsupport TensorFlow
Tools: PLAI modeltraining tool(Support for GG1, GNet18 and GNetfc network models based onVGG-16)
OS: SupportUbuntu, Windows operating system
Size : 27.5x12.5x3.5mm
Shipping List:
NCCS1 Neural Network Computing Card x1
Screwsx1
Technical Support:
Wiki:http://wiki.t-firefly.com/en/NCCS1/
Expert Support:firefly@ smartfire.cn Skype:smartfire_cn
Product Customization:firefly@smartfire.cn
