Accelerate AI with hardware

After many years of processing AI workloads in the cloud companies are moving their workloads off the cloud. The most common reasons:

Total cost of ownership

Running a GPU-powered virtual machine in any cloud costs more than 10,000 USD yearly. Comparing that to a traditional investment in a server with next-day business support and full waranty, in many cases the total cost of ownership is higher and more unpredictable with cloud solutions.


Due to sensitive data or regulatory compliance it might not be possible to move data into a cloud requiring AI acceleration hardware to run in local environments.

Existing infrastructure

Lastly, many companies have made long-term investments in on-premise IT and have no tooling or expertise in hybrid-cloud solutions. In such cases AI acceleration hardware is ideal to bring data science into the existing infrastructure.

Huawei Hardware Acceleration

Huawei are global leaders in cost-effective hardware acceleration benefitting companies with existing IT on-premise infrastructure and companies looking to reduce their cloud expenses on AI.

Atlas 200 Developer Kit
Huawei Logo

Atlas 200 Developer Kit


Powered by the Huawei Ascend 310 processor, the Atlas 200 DK AI Developer Kit helps AI application developers quickly get familiar with the development environment. The device provides external ports for developers to quickly and easily access and use the powerful processing capability of the Ascend 310 processor. Equipped with the Atlas 200 AI accelerator module as its core component, the Atlas 200 DK AI Developer Kit provides the main service ports of the Ascend 310 processor on the expansion baseboard through a high-speed connector. Thanks to the full-stack, all-scenario capability of the Ascend processor, programs can be developed on the Atlas 200 DK AI Developer Kit in a one-off manner and deployed in device-edge-cloud scenarios with zero code modification. It can be used in various fields such as safe cities, drones, robots, video servers, and smart gates. #16TOPS@INT8 #8GB LPDDR4x 32.9 mm x 137.8 mm x 93.0 mm

Order Form