Deepen the developer ecosystem and accelerate the innovative development of the AI ​​industry

Intel and many partners gather at 2022 OpenVINO™ DevCon

June 27, 2022, Shanghai - Recently, the 2022 OpenVINO™ DevCon•China Station hosted by Intel was successfully held in the form of an online summit, with many leading companies from all walks of life gathered here. At this global OpenVINO™ developer event, OpenVINO™ core technology experts shared the latest product information and technical demonstrations, and many experts also shared and communicated in multiple dimensions such as industry insights, core technologies, and case presentations to help developers improve their technical level while broadening their professional horizons and understanding industry trends. This conference also has a "Developer Ecosystem AMA (Ask Me Anything)" session to support developers and professional technical experts to have in-depth exchanges and active interactions, and to escort AI practitioners.

OpenVINO™ DevCon is a global event dedicated to encouraging OpenVINO™ developers to improve their skills and exchange knowledge. 2022 is the first year of this global event, which will start in the United States on June 13 and will be held in China, Japan and other countries and regions . On June 25, the first meeting of the OpenVINO™ DevCon series of summits in China was held, combining local Chinese cases, aiming to better help developers use OpenVINO™ to improve their skills, broaden their career horizons and gain insight into the latest industry trends.

“If we look at what OpenVINO™ has been like over the past four years, it’s grown a lot from when we first started,” said Adam Burns, general manager of OpenVINO™ developer tools in the Network and Edge Group at Intel. “As the product has grown, so has our community. So I think now is the perfect time to have DevCon, not only to showcase the product updates that we’ve been working on, but also to really provide an opportunity for everyone to connect with partners and see some great innovations coming out of the OpenVINO™ developer community.”

Since the launch of OpenVINO™ in 2018, Intel has helped hundreds of thousands of developers significantly improve AI inference performance and expand their applications from intelligent edge computing to cloud computing and server-side. At the end of the first quarter of 2022, Intel launched a new version of the distribution OpenVINO™ toolkit. The new features are mainly developed based on developer feedback over the past three and a half years, including easier update APIs, a wider selection of deep learning models, more device portability options, and higher inference performance and less code configuration, designed to simplify and automate model deployment. The technical highlights of the new version of OpenVINO™ 2022.1 include: automatically balancing the inference task load between CPUs, GPUs, and other accelerators; more conveniently importing pre-trained models from other frameworks; automatically optimizing and parallelizing inference tasks; improving NLP-related network performance through the Dynamic Shape feature; and optimizing models during or after training.

At the summit, Intel technical experts also shared two technical demonstrations on the technical highlights of the new version of OpenVINO™:

• AUTO Device reduces initialization waiting time and optimizes performance indicators: Among the many functions and configuration options that OpenVINO™ has provided, developers can deeply optimize the performance of the corresponding underlying devices for their own applications. As the application field expands, OpenVINO™ has launched a new AUTO virtual device to help developers more easily develop general programs that can be adapted to different platforms and have good performance.

• OpenVINO™ support for NLP and dynamic input: In NLP machine question-answering scenarios, use OpenVINO™ for fast reasoning; in optical character recognition (OCR) scenarios, use OpenVINO™ to reason about PaddleOCR, directly read the model based on the PaddlePaddle deep learning framework, and complete real-time AI reasoning for the PaddleOCR model with dynamic input.

In addition to the key technical sharing of the new version of OpenVINO™ 2022.1 , several Intel ecosystem partners including Baidu PaddlePaddle, Shanghai Trend Information Technology Co., Ltd., YiQi Education Technology and Fudan University also appeared at the summit and shared how they promoted the innovative development of the AI ​​industry through OpenVINO™ and co-created a developer ecosystem with Intel. Dr. Wei Chen, Vice President of Intel's Network and Edge Division and General Manager of the Video Division, said: "For a long time, we have been committed to working with many ecosystem partners to provide developers with a good learning platform and project practice opportunities. Based on this, we have not only launched the OpenVINO™ toolkit worldwide, but also launched the Intel® Edge Software Center and Intel® DevCloud, so that global customers can make full use of the OpenVINO™ toolkit and support developers to easily access tools and technologies anytime, anywhere. Now, we are gradually integrating these toolkits with other assets of the Network and Edge Division (such as SmartEdge, DPDK and other new technologies) to provide developers with unprecedented tools to help them complete their work more smoothly."

In addition, with the rapid development of the AI ​​industry, facing the increasing market demand for AI professionals, Intel is also continuously injecting new momentum into the training of AI professionals and helping developers grow. Intel has a complete developer training program. Through cooperation with educational institutions, universities, and industry ecosystem partners, it provides developers with a wealth of courses to learn and meet the needs of all stages, so that everyone can go from entry to proficiency instead of from entry to giving up. In addition, Intel is building a healthy developer ecosystem, disseminating knowledge and skills in various developer communities, and connecting everyone through activities to focus on the future of AI.

Intel has always believed that developers are the core driving force behind the innovative development of the AI ​​industry. For a long time, Intel has continued to work hard to build a developer ecosystem, and in the process, it has collected developer feedback, promoted communication between developers, and encouraged developers to participate in code contributions, thereby helping developers learn and improve their professional level. In the future, Intel will continue to work with industry partners to help more developers achieve self-growth and promote the large-scale implementation of AI applications.