Seamless Migration, Quicker Time to Market​

To eliminate painstaking challenges involved in migrating trained AI models into production, Deep Vision offers a highly optimized set of automated development tools. The software and processors are orchestrated together to support implementation of complex, streamlined AI applications.

  • Effortless deployment of AI models on Silicon
  • Extensible Architecture to address future models​
  • Scalable performance supports application needs​

Ease of Use

Ease of Use

Friction-free porting of trained AI models to the Deep Vision ARA-1 Edge AI processor​

LEARN MORE

Accuracy

Performance

Unmatched accuracy using proprietary quantization algorithms

LEARN MORE

Performance

Accuracy

Intelligently utilize optimal cores and scale to fit the needs of any Deep Learning application

LEARN MORE

Frameworks

Frameworks

Support for popular frameworks including Caffe, TensorFlow, PyTorch, MXNet and ONNX

LEARN MORE
Seamless Migration, Quicker Time to Market Seamless Migration, Quicker Time to Market Seamless Migration, Quicker Time to Market Seamless Migration, Quicker Time to Market

Uncompromised Compute in a Small Form Factor​

The Deep Vision ARA-1 Edge AI processor offers computational flexibility, high programmability, and low power, combined with the integrated resource management and software necessary to put computationally complex AI-based vision applications rapidly into production.

Uncompromised Compute in a Small Form Factor
Unparalleled Results

Unparalleled Results

Deep Vision’s patented Polymorphic Dataflow Architecture optimizes implementation of each layer in a neural network to deliver high performance and power efficiency, while enabling incredible flexibility and control over the execution of complex AI inference applications. Micro-dataflow patterns in compute cores support data expansion and reduction (map/reduce) resulting in minimal data movement between compute and memory while supporting a variety of dataflows at the tool suite level.

LEARN MORE

One Processor, Endless Possibilities​

Deep Vision untethers the complexities to enable the creation of new world AI applications at the edge. Multiple industries now have the power to bring high-performance processing to more devices, creating new solutions in retail, manufacturing, automotive, security, and more.

LEARN MORE
One Processor, Endless Possibilities​