This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search

Reminder

To have a better experience, please upgrade your IE browser.

upgrade
If you need help, please click here:

Customer Background

As a leading technological company, SenseTime focuses on computer vision and deep learning. With the first-class AI algorithms, SenseTime commands the largest market share in multiple vertical fields.

Challenges

Autonomous driving research needs to analyze and compute massive data obtained from cameras. Data storage, conversion, and model training require a hardware system with high computing performance, throughput, and reliability.

The SenseTime autonomous driving research institute in Japan has its own algorithms and is eagerly seeking suitable hardware to support new algorithm research. With the advancement of research, data volumes and analytics performance change dynamically. SenseTime expects the AI computing resources to be flexibly expanded as required.

Solution

Huawei has a deep collaboration relationship with SenseTime. SenseTime also has a deep accumulation in autonomous driving technology. SenseTime selects Huawei's Atlas AI servers as the infrastructure of the L4 autonomous driving solution.

The Huawei Atlas AI server adopts a modular design to implement flexible AI computing resource expansion and provide one-click topology switching to meet the performance requirements of different application models. A single Atlas AI server supports 8 AI training acceleration cards, meeting the performance requirements of AI training for autonomous driving.

Benefits

Using Huawei Atlas AI servers, SenseTime quickly sets up a large-scale autonomous driving training cluster with a small amount of investment. The cluster reduces the CAPEX and meets the requirements for fast capacity expansion of AI computing resources in the future.

The Huawei Atlas AI server supports one-click topology switching to quickly adapt to different AI models, improving system resource utilization and reducing OPEX.

0 readers

(0 scores)

Like the story? Give your score.

0/500

Write your comment here.
Submit

0 comments

    More comments

      You have scored successfully.

      You have submitted successfully.

      Evaluation failed.

      Submission failed.

      Please write your comment first.