OpenVINO 2023.1 Released - More GenAI, Expanded LLM Support & Meteor Lake VPU

Intel's OpenVINO 2023.1 was just published to GitHub as the newest version of this open-source toolkit for optimizing and deploying AI workloads across their CPUs, GPUs, and now also having official support for the new VPU being found with Meteor Lake SoCs.

OpenVINO 2023.1 brings new Generative AI "GenAI" options with further enhancing its PyTorch model support and also working on optimized support with Hugging Face. OpenVINO 2023.1 also has broader large language model (LLM) support to yield better runtime performance and at reduced memory use, improved LLM GPU support, and a new 8-bit weights compression method for the Neural Network Compression Framework.

The OpenVINO 2023.1 release also has support for the VPU found with forthcoming Meteor Lake / Intel Core Ultra SoCs. This should further enhance the OpenVINO AI workload performance on upcoming Intel laptops. This new OpenVINO release also brings MediaPipe integration for greater access for building multi-purpose AI pipelines.

With the Intel VPU abound with Meteor Lake and already some Linux driver work showing an improved version with Lunar Lake, OpenVINO 2023.1 is deprecating its Intel Gaussian & Neural Accelerator (Intel GNA) support with that hardware block found on SoCs going back to the Tiger Lake days. Intel GNA is now superseded by the Intel VPU and the OpenVINO software integration will be dropped next year.

Downloads and more details on today's OpenVINO 2023.1 release can be found via OpenVINO on GitHub.