Beijing Plink AI Technology Co., Ltd
                                                                                                           
Verified Supplier
17 Years
Since 2009
Menu

Internet Atlas 300I Inference Card Speech Analysis

Price Negotiable
Price: To be negotiated
MOQ: 1set
Delivery Time: 15-30 work days
Brand: Hua Wei
Product Description

 

Internet Ascend Atlas 300I Duo Inference Card Speech analysis 280 TOPS 48GB Memory video applications

 

Hua Wei Ascend Atlas 300I Duo Inference Card

The Hua Wei Ascend Atlas 300I Duo Inference Card integrates the general-purpose processor, AI Core, and codec to provide powerful AI inference and video analysis functions. With advantages such as superior computing power, ultra-high energy efficiency, and high-performance video analysis, it is suitable for a wide array of scenarios such as Internet, smart city, and smart transportation, and content analysis, OCR, speech recognition, and video analysis. The typical scenarios of Hua Wei Ascend Atlas 300I Duo inference card include search recommendation, content review, OCR system, target recognition system, and Smart City 2.0 city governance.Based on user inputs (such as user profiles and search keywords), a search recommendation system uses the recall and sorting algorithms to filter out the recommended materials (such as videos and text) from the content pool. The system is mainly used in the Internet and many other fields.

 

Specifications of the Hua Wei Ascend Atlas 300I Duo Inference Card

Form factor

FHFL single-slot (10.5 inches)

AI computing power

Half-precision (FP16): 140 TFLOPS

Integer precision (INT8): 280 TOPS

Memory

LPDDR4X

Capacity: 48 GB

Total bandwidth: 408 GB/s

Error checking and correcting (ECC)

Encoding/ Decoding capability

H.264/H.265 decoder hardware decoding, 256-channel 1080p 30 FPS or 32-channel 4K 60 FPS

 

H.264/H.265 encoder hardware encoding, 48-channel 1080p 30 FPS or 6-channel 4K 60 FPS

 

 JPEG decoding capability of 4K 1024 FPS and encoding capability of 4K 512 FPS, maximum resolution of 8192 x 8192

Virtual instance specifications

One Ascend AI Processor can be divided into several virtual NPUs in virtualization mode. Each virtual NPU supports 1, 2, or 4 AI Cores, and other hardware resources (such as memory and codec module) are divided proportionally.

Power consumption

Maximum power consumption: 150 W

Dimensions (W x H x D)

266.7 mm × 18.46 mm × 111.15 mm (10.5 in. x 0.72 in. x 4.37 in.)

Weight (g)

910 g

PCIe port

x16 lanes, compatible with x8/x4/x2 PCIe Gen4.0, compatible with 3.0/2.0/1.0

Operating Temperature

0°C to 55°C (32°F to 131°F)

 

Hua Wei Ascend Atlas 300I Duo Inference Card

Heat Dissipation Specifications

The Hua Wei Ascend Atlas 300I Duo inference card supports an air intake vent temperature from 0°C to 55°C. Internal temperature monitoring points are available. Both in-band and out-of-band modes can be used to monitor the Ascend 310P and storage chip in real time. To ensure heat dissipation, the temperature of the card must be lower than the specification limits when it is working

Item

Ascend 310P Temperature(°C)

Storage Chip Temperature(°C)

Power-off temperature

 

106 100

Underclocking temperature

103 90

Long-term operating temperature

102 85

 

Hua Wei Ascend Atlas 300I Duo inference card Application Scenarios

 

Integrated in servers for AI inference

Get in Touch

Have questions about our products or want to discuss a custom order? Our team is ready to help you.

Company Beijing Plink AI Technology Co., Ltd
Location C1106,Jinyu Jiahua Building,Shangdi 3rd Street, Haidian District, Beijing 100085, P.R.China
Contact Person Cindy

Request A Quote

Please check your email address.
Your message must be at least 20 characters.