May 20, 2022, Beijing, China - Today, Graphcore® (to be announced) officially announced its participation in the hardware ecosystem co-creation plan at the Wave Summit 2022 Deep Learning Developer Summit . Based on the co-creation plan, Graphcore and Baidu Paddle will jointly develop technical solutions, jointly customize the Paddle framework, build model libraries and scenario examples, empower the industry with "IPU+Paddle", and promote the AI transformation and upgrading of the industry. At present, Poplar® SDK 2.3 and Baidu Paddle 2.3 have been fully integrated, and the relevant code will be available on Baidu Paddle's GitHub today for developers to obtain.
Baidu PaddlePaddle is China's first independently developed, feature-rich, open-source, industrial-grade deep learning platform. As of May 2022, Baidu PaddlePaddle has gathered 4.77 million developers, created 560,000 AI models, and served 180,000 enterprises and institutions. The hardware ecosystem co-creation plan is an important part of the PaddlePaddle "Great Voyage" Plan 2.0. As one of the participating members of the plan , Graphcore will work with Baidu PaddlePaddle to further build the IPU-based Baidu PaddlePaddle framework, enrich the model library and scenario examples for various AI application scenarios such as cloud and data centers, Internet, finance, medical and life sciences, automobiles, education, smart retail, etc., and use IPU to accelerate the innovation and application of AI technology in various fields.
Graphcore is one of the initial members of the Baidu Paddle hardware ecosystem and an important hardware partner of Baidu Paddle. In 2021, Baidu Paddle has achieved full support for training and reasoning on Graphcore IPU and open sourced the relevant code. The two teams have achieved good performance results in data parallelism and model parallelism on IPU-POD16, as well as intensive reading and throughput verification on the Bert-Base model. The results show that the Baidu Paddle solution running on IPU hardware can achieve good acceleration effects on current mainstream AI training tasks and cutting-edge AI innovative model tasks. Recently, Graphcore released the Bow system with the world's first 3D WoW processor Bow IPU as the core. Compared with the previous generation of products, the Bow system can provide a 40% performance improvement and a 16% power efficiency improvement without changing the existing code. Baidu Paddle developers can migrate from the previous generation IPU system to the latest Bow system without changing the existing code, and accelerate the development process with the powerful performance of the Bow system.
Jin Chen, Vice President of Engineering at Graphcore China and AI algorithm scientist, said: "As an excellent industrial-grade deep learning platform, Baidu PaddlePaddle has been favored and widely adopted by developers. While deepening its own field, it fully plays the role of the platform and works with partners to build an ecosystem, providing a foundation for co-creating tools, models, industry cases and practical experience. Graphcore is very pleased to continue to deepen its cooperation with Baidu PaddlePaddle, give full play to the acceleration role of IPU, allow efficient AI computing power to flow smoothly, provide computing support for Baidu PaddlePaddle developers, and work together to enable the transformation and upgrading of industrial AI in various fields."
百度飞桨产品团队负责人赵乔表示:“Graphcore是百度飞桨长期以来的重要合作伙伴。Graphcore的IPU系统和Poplar软件在我们双方的合作中展现出了令人印象深刻的易用性、灵活性和可扩展性,在自然语言处理、计算机视觉、语音等AI领域都展示出了显著的加速效果,在许多应用场景都展现出了巨大的应用潜力。我们期待在硬件生态共创计划中与Graphcore进一步深化合作,共同推进AI技术的应用,以AI赋能各产业的创新和发展。”
In the future, Graphcore and Baidu PaddlePaddle plan to explore the application of Baidu PaddlePaddle on the Bow system, and through continuous technological upgrades and continuous optimization of functional experience, accelerate the implementation of the AI industry and achieve ecological prosperity and win-win results.
Previous article:Intel and Baidu PaddlePaddle join hands to create an integrated hardware and software artificial intelligence ecosystem
Next article:Microsoft enters the "industrial metaverse" Kawasaki will use HoloLens devices to manufacture robots
- Popular Resources
- Popular amplifiers
- CGD and Qorvo to jointly revolutionize motor control solutions
- Advanced gameplay, Harting takes your PCB board connection to a new level!
- Nidec Intelligent Motion is the first to launch an electric clutch ECU for two-wheeled vehicles
- Bosch and Tsinghua University renew cooperation agreement on artificial intelligence research to jointly promote the development of artificial intelligence in the industrial field
- GigaDevice unveils new MCU products, deeply unlocking industrial application scenarios with diversified products and solutions
- Advantech: Investing in Edge AI Innovation to Drive an Intelligent Future
- CGD and QORVO will revolutionize motor control solutions
- Focusing on promoting innovation and transformation, UHV field creates more possibilities in the era of clean energy
- Infineon Technologies Launches SECORA™ Pay Green, World’s First Contactless Payment Card Technology, Reducing Plastic Waste by Up to 100%
- LED chemical incompatibility test to see which chemicals LEDs can be used with
- Application of ARM9 hardware coprocessor on WinCE embedded motherboard
- What are the key points for selecting rotor flowmeter?
- LM317 high power charger circuit
- A brief analysis of Embest's application and development of embedded medical devices
- Single-phase RC protection circuit
- stm32 PVD programmable voltage monitor
- Introduction and measurement of edge trigger and level trigger of 51 single chip microcomputer
- Improved design of Linux system software shell protection technology
- What to do if the ABB robot protection device stops
- Detailed explanation of intelligent car body perception system
- How to solve the problem that the servo drive is not enabled
- Why does the servo drive not power on?
- What point should I connect to when the servo is turned on?
- How to turn on the internal enable of Panasonic servo drive?
- What is the rigidity setting of Panasonic servo drive?
- How to change the inertia ratio of Panasonic servo drive
- What is the inertia ratio of the servo motor?
- Is it better for the motor to have a large or small moment of inertia?
- What is the difference between low inertia and high inertia of servo motors?
- WIFI6 Explained
- Control battery output and prevent battery reverse connection
- 【NXP Rapid IoT Review】 + Kit Modification-External Lithium Battery (Link)
- How to design a typical analog front-end circuit
- [Fudan Micro FM33LG0 Series Development Board Review] Driving RGB TFT Display
- Thank you for being there + thank you for everyone who appears in my life
- Newbies need help with a few questions about RA2A1
- The virtual machine started to freeze again
- How to solve the problem of solder balls caused by mismatched pads during production
- How speakers work