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


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

If you need help, please click here:

Atlas 300 AI Accelerator Card

Based on the Ascend 310 AI processor, the Atlas 300 AI accelerator card is packaged in a standard PCIe card form factor. It supports multiple types of data precision and is suitable for AI training and inference scenarios. The Atlas 300 AI accelerator card uses a groundbreaking Da Vinci architecture to unlock industry-leading performance and energy efficiency. It can be widely used for data centers and the intelligent edge.

Application Scenarios

  • Safe City

    For facial recognition in scenarios such as public security management, customs and border defense, safe campus, and population database cross-check.

  • Smart Transportation

    For traffic violation investigation and vehicle tracking in applications such as license plate recognition, vehicle big data analytics, and track analysis.

  • Smart Finance

    For automatic identification and content of financial documents, financial statements, cards, and insurance claims; supports service processes on smartphones.



Model Atlas 300 AI Accelerator Card
Form Factor Half-height, half-length PCIe standard card
Memory 4 x LPDDR, 32 GB, and 3,200 Mbit/s
AI Compute Power 64 TOPS INT8
Encoding/Decoding Capability • H.264 hardware decoding, 64-channel 1080p30 (2-channel 3,840 x 2,160 @ 60 FPS)
• H.265 hardware decoding, 64-channel 1080p30 (2-channel 3,840 x 2,160 @ 60 FPS)
• H.264 hardware encoding, 4-channel 1080p30
• H.265 hardware encoding, 4-channel 1080p30
• JPEG decoding, 4-channel 1080p256; encoding 4-channel 1080p64
• PNG decoding, 4-channel 1080p48
PCIe Expansion • PCIe 3.0, compatible with 2.0/1.0
• x16 lanes, compatible with x8/x4/x2/x1
Power Consumption 67W
Dimensions (W x D) 169.5 mm x 68.9 mm (6.67 in. x 2.71 in.)
Weight 319g
Operating Temperature 0°C to 55°C (32°F to +131°F)
Model Atlas 300 AI Accelerator Card
Form Factor standard full-height 3/4-length PCIe card
Memory built-in 32 GB HBM and 16 GB two-level, large-capacity memory
AI Compute Power 256T FLOPS FP16
PCIe Expansion PCIe 4.0x16

For Partners

If you are already a Partner, please click here to get more marketing resources.
Click here to visit the Partner zone to check enquiry status, manage orders, get support, or learn more about Huawei Partners.