Optimizing Inference with Support for the OpenVINO™ Toolkit

Modzy now supports models prepared, optimized, and compressed with the OpenVINO toolkit, while also enabling users to take advantage of accelerated inference with the OpenVINO runtime. This integration makes it easy to convert and optimize models from popular training frameworks to deploy and run on a mix of Intel hardware and environments, on-prem, in the cloud, or at the edge.  

The Modzy platform provides powerful AI inferencing in the cloud, at the edge, and everywhere in between. With support for OpenVINO optimization, compression, and runtime, Modzy customers can now unlock huge performance gains when running AI models on Intel processors for many use cases and solutions across a range of industries, especially those that require AI insights at the edge.

Accelerated Inference with OpenVINO

In our tests, we found that running a ResNet model converted to the OpenVINO IR format with OpenVINO runtime on Modzy results in up to a 9x performance increase compared to running the model with Torch. This performance increase makes it easier than ever for our customers to launch real-time computer vision applications on Intel processors at the edge and in the cloud.

Industries and Use Cases

Because Modzy provides the ability to run OpenVINO accelerated models in the cloud, on-prem, or at the edge, these models can be integrated into solutions across a range of industries.  

Manufacturing Many manufacturing use cases require real-time AI insights delivered at the edge with minimal latency. The powerful combination of Modzy plus OpenVINO accelerated models can be used to develop solutions to drive efficiency, improve product quality, and ensure safety throughout a manufacturing facility.  

Telecommunications AI can be used for fleet management, asset tracking, IoT solutions, network optimization, predictive maintenance, and to improve customer experience for telecommunications providers. With Modzy’s ability to deploy, run, and orchestrate OpenVINO accelerated models in a disconnected environment, customers can develop solutions that yield more accurate insights to improve operations, reduce costs, and increase revenue.

Smart Cities By integrating computer vision for video analytics solutions, cities can monitor activities in real-time to improve safety and security. Using Modzy to run OpenVINO optimized models can ensure the inference acceleration needed to enable real-time AI generated insights at scale.

Getting Started with OpenVINO on Modzy

To get started deploying and running OpenVINO accelerated models on Modzy, contact info@modzy.com.

‍