Enquiries
This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>
Atlas 200 AI Accelerator Module
The Atlas 200 AI Accelerator Module (model: 3000) integrates the Ascend 310 AI processor to implement facial recognition and image classification on the device side. It is widely used in AI scenarios such as the intelligent camera, robot, and drone.
• Provides 16 TOPS INT8 computing power and supports 16-channel real-time HD video analytics in the size of half a credit card.
• Supports 2 computing power configurations: 16 TOPS and 8 TOPS.
• Supports milliwatt-level hibernation and wake-up in milliseconds, enabling edge AI applications with the power consumption as low as 9.5 W.
Parameters | Model |
Model | Atlas 200 AI Accelerator Module Model: 3000 |
AI Chip | Ascend 310 |
Memory | 4/8 GB LPDDR4X, total bandwidth 51.2 Gbit/s |
AI Compute Power | 16 TOPS INT8 8 TOPS INT8 |
Encoding/Decoding Capability | • H.264 hardware decoding, 16-channel 1080p 30 FPS (2-channel 3840 x 2160 @ 60 FPS) • H.265 hardware decoding, 16-channel 1080p 30 FPS (2-channel 3840 x 2160 @ 60 FPS) • H.264 hardware encoding, 1-channel 1080p 30 FPS • H.265 hardware encoding, 1-channel 1080p 30 FPS • JPEG decoding: 1080p 256 FPS; encoding: 1080p 64 FPS; maximum resolution: 8192 x 4320 • PNG decoding: 1080p 24 FPS; maximum resolution: 4096 x 2160 |
Port | • 4 PCIe 3.0 • 1 USB2.0/USB3.0 • 1 RGMII |
Serial Bus | UART/I2C/SPI |
Interface Specifications | 144-pin BTB connector |
Power Consumption | Max. 9.5 W |
Operating Temperature | -25°C to +80°C (-13°F to +176°F) |
Dimensions (H x W x D) | 8.5 mm x 52.6 mm x 38.5 mm (0.33 in. x 2.07 in. x 1.52 in.) |
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.