Huawei Unveils AI Strategy and Full-Stack, All-Scenario AI Portfolio

Huawei Unveils AI Strategy and Full-Stack, All-Scenario AI Portfolio

Huawei the Chinese manufacturer conducted the third annual HUAWEI CONNECT at Shanghai which is still live (October 10- October 12) and the cornerstone of the event is the “Activate Intelligence” and its challenges, opportunities, innovations, and practices. The company believes that by the year 2025, there will more the 40 billion smart devices across the globe and 90% of the users will be using the smart digital assistant.

Ten future changes: Driving Huawei’s AI strategy

Proactive change is the first step towards a better future in AI. Huawei has defined ten changes that will help pave the way. They include:

  • Faster model training
  • Abundant and affordable computing power
  • AI deployment and user privacy
  • New algorithms
  • AI automation
  • Practical application
  • Real-time, closed-loop system
  • Multi-tech synergy
  • Platform support
  • Talent availability
“Huawei’s AI strategy is to invest in basic research and talent development, build a full-stack, all-scenario AI portfolio, and foster an open global ecosystem,” said Eric Xu during his keynote.

Keeping in mind the above points, the company at the event announced two chips- The Ascend 910 and Ascend 310.

The Ascend 910 will be based on the 7nm technology and 350W maximum power consumption to offer faster processing of data in less time as compared to its competitors, the company said. The Ascend 910 will be launched next year in the second quarter. The Ascend 310 will also be launched later focused on edge computing devices such as the Internet of Things devices and mobile phones.

Huawei Unveils AI Strategy and Full-Stack, All-Scenario AI Portfolio

Apart from the Ascend series of chips, Huawei’s full-stack AI portfolio also includes the following:

  • CANN (Compute Architecture for Neural Networks): A chip operators library and highly automated operators development toolkit
  • MindSpore: A unified training and inference framework for device, edge, and cloud (both standalone and cooperative)
  • Application enablement: Full-pipeline services (ModelArts), hierarchical APIs, and pre-integrated solutions.

LEAVE A REPLY

Please enter your comment!
Please enter your name here