CEVA cores look further afield
- 著者:Ella Cai
- 公開：:2018-01-08
In Q2, CEVA will start selected licensing of a processor family called NeuPro ‘for deep learning inference at the edge’ with performance ranging from 2-12.5 TOPS.
General licensing starts in Q3.
The NeuPro processor line goes beyond machine vision to applications like natural language processing, real-time translation, authentication, workflow management, and other learning-based applications.
“It’s abundantly clear that AI applications are trending toward processing at the edge, rather than relying on services from the cloud. ‘The computational power required along with the low power constraints for edge processing, calls for specialized processors rather than using CPUs, GPUs or DSPs,” says CEVA’s Ilan Yona.
The NeuPro architecture is composed of a combination of hardware-based and software-based engines coupled for a scalable and expandable solution. Optimizations for power, performance, and area (PPA) are achieved using a precise mix of hardware, software and configurable performance options for each application tier.
The NeuPro family comprises four AI processors offering different levels of parallel processing:
· NP500 is the smallest processor, including 512 MAC units and targeting IoT, wearables and cameras
· NP1000 includes 1024 MAC units and targets mid-range smartphones, ADAS, industrial applications and AR/VR headsets
· NP2000 includes 2048 MAC units and targets high-end smartphones, surveillance, robots and drones
· NP4000 includes 4096 MAC units for high-performance edge processing in enterprise surveillance and autonomous driving
Each processor consists of the NeuPro engine and the NeuPro VPU. The NeuPro engine includes the hardwired implementation of neural network layers among which are convolutional, fully-connected, pooling, and activation.
The NeuPro VPU is a cost-efficient programmable vector DSP, which handles the CDNN software and provides software-based support for new advances in AI workloads. NeuPro supports both 8-bit and 16-bit neural networks, with an optimized decision made in real time in order to deliver the best tradeoff between precision and performance.
The MAC units achieve better than 90% utilization when running, ensuring highly optimized neural network performance. The overall processor design reduces DDR bandwidth substantially, improving power consumption levels for any AI application.
The NeuPro family, coupled with CDNN, CEVA’s neural network software framework, provides the a way for developers to generate and port their proprietary neural networks to the processor. CDNN supports the full gamut of layer types and network topologies, enabling fastest time-to-market.
In conjunction with the NeuPro processor line, CEVA will also offer the NeuPro hardware engine as a Convolutional Neural Network (CNN) accelerator. When combined with the CEVA-XM4 or CEVA-XM6 vision platforms, this provides a flexible option for customers seeking a single unified platform for imaging, computer vision and neural network workloads.
General licensing starts in Q3.
The NeuPro processor line goes beyond machine vision to applications like natural language processing, real-time translation, authentication, workflow management, and other learning-based applications.
“It’s abundantly clear that AI applications are trending toward processing at the edge, rather than relying on services from the cloud. ‘The computational power required along with the low power constraints for edge processing, calls for specialized processors rather than using CPUs, GPUs or DSPs,” says CEVA’s Ilan Yona.
The NeuPro architecture is composed of a combination of hardware-based and software-based engines coupled for a scalable and expandable solution. Optimizations for power, performance, and area (PPA) are achieved using a precise mix of hardware, software and configurable performance options for each application tier.
The NeuPro family comprises four AI processors offering different levels of parallel processing:
· NP500 is the smallest processor, including 512 MAC units and targeting IoT, wearables and cameras
· NP1000 includes 1024 MAC units and targets mid-range smartphones, ADAS, industrial applications and AR/VR headsets
· NP2000 includes 2048 MAC units and targets high-end smartphones, surveillance, robots and drones
· NP4000 includes 4096 MAC units for high-performance edge processing in enterprise surveillance and autonomous driving
Each processor consists of the NeuPro engine and the NeuPro VPU. The NeuPro engine includes the hardwired implementation of neural network layers among which are convolutional, fully-connected, pooling, and activation.
The NeuPro VPU is a cost-efficient programmable vector DSP, which handles the CDNN software and provides software-based support for new advances in AI workloads. NeuPro supports both 8-bit and 16-bit neural networks, with an optimized decision made in real time in order to deliver the best tradeoff between precision and performance.
The MAC units achieve better than 90% utilization when running, ensuring highly optimized neural network performance. The overall processor design reduces DDR bandwidth substantially, improving power consumption levels for any AI application.
The NeuPro family, coupled with CDNN, CEVA’s neural network software framework, provides the a way for developers to generate and port their proprietary neural networks to the processor. CDNN supports the full gamut of layer types and network topologies, enabling fastest time-to-market.
In conjunction with the NeuPro processor line, CEVA will also offer the NeuPro hardware engine as a Convolutional Neural Network (CNN) accelerator. When combined with the CEVA-XM4 or CEVA-XM6 vision platforms, this provides a flexible option for customers seeking a single unified platform for imaging, computer vision and neural network workloads.
