Huawei Ascend B300/B300E
Huawei Ascend B300/B300E
1
2
3

Huawei Ascend B300/B300E

Huawei Ascend B300/B300E with Atlas 200 DK Acceleration Module is a high-performance AI application edge computer, boasting 20 TOPS AI computing power. Proudly produced domestically, it features an all-aluminum alloy body and fanless passive cooling design, ensuring low power consumption and stable operation in high-intensity environments. Ideal for intelligent cameras, robots, industrial control systems, medical equipment, drones, and more, it enables efficient target recognition and image classification at the edge, delivering reliable performance across diverse AI-driven applications.

Technical Reference

Product Details

Accessories

Atlas 200I DK A2 Developer Kit
AI Processor20 TOPS disposition： 4 core * 1.6 GHz8 TOPS disposition： 4 core * 1.0 GHz
Codec capability20 TOPS Configuration:
Support H.264 / H.265 hardware decoding, 40 channels 1080P 30FPS, 4 channels 4K (3840 x 2160) 75FPS;
Support H.264 / H.265 hardware coding, 20 channels 1080P 30FPS, 3 channels 4K (3840 x 2160) 50FPS;
8 TOPS Configuration:
Support H.264 / H.265 hardware decoding, 20 channels 1080P 30FPS, 2 channels 4K (3840 x 2160) 75FPS;
Support H.264 / H.265 hardware coding, 12 channels 1080P 30FPS, 2 channels 4K (3840 x 2160) 50FPS;
20/8 TOPS Configuration:
JPEG decoding capacity 1080P 512FPS, encoding capacity 1080P 256FPS, maximum resolution: 16384 x 16384;
AI computing power20 TOPS INT8, 10 TFLOPS FP16；
8 TOPS INT8, 4 TFLOPS FP16；
RAM（Internal memory）Type: LPDDR4X;
Speed: 4266Mbps;
Support ECC;
Memory: 12GB;
Bit width: 96bit.
ROM（Built-in storage capacity）Standard eMMC,32G Byte, (32G/64G/128G optional) Maximum 256G Byte;
Built-in SPI flash, capacity 64MB;
Support external MMC interface, can support: - eMMC5.1 particles, support the highest mode HS400;
Support for external SATA interfaces - up to 4 SATA 3.0 interfaces, backward compatible with SATA 2.0 and SATA 1.0, SATA PM (SATA Port Multiplier);
Support TF card storage;
USB3.0 port2 standard USB3.0-A external interfaces (5Gbps);
One onboard expandable USB3.0 interface;
TYPE C portUSB3.0 type-c Device Indicates one interface
Gigabit network interfaceTwo 1000BASE-R (1.25Gpbs) network interfaces
M.2-key-M slotSupports the M.2-key-M slot of NVMe SSD 2280;
SATA 3.0 (6Gbps), backward compatible SATA 2.0 (3Gbps) and SATA 1.0 (1.5Gbps), PCIe Gen3 (8Gbps) backward compatible PCIe Gen2 (5Gbps) and PCIe Gen1 (2.5Gbps) Flexible configuration;
M.2-key-B slot5G module expansion slot (4G/5G can only choose one)
(support 3 major operators: mobile, Unicom, telecom)
mini-PCIE slot4G module expansion slot (4G/5G can only choose one)
(Support 3 major operators: mobile, Unicom, telecom).
M.2-key-E slotSupport 2.4+5GHz WiFi, support Wi-Fi 802.11b/g/n/ac protocol;
Support Bluetooth function, V2.1+EDR/Bluetooth 3.0/3.0+HS/4.2/BLE;
TF slotExternal card Standard TF card
SIM card slotExternal Board Micro SIM Card (USIM)
Video output interfaceHDMI: 2 off-board outputs;
In-board output MIPI-DSI: 4Lane;
Video input interfaceThe Atlas 200I A2 module comes with an on-board 2* MIPI-CSI Camera function, RaspberryPi Camera V2
Analog audio inputOff-board analog audio input 1 way
Fan interfaceOne fan port on the board
RTC battery portReal-time clock power battery interface
powerOne
Software upgrade keyOne
Reset keyOne
Dip switch4-bit DIP switch, 3-bit for boot boot mode, and 1-bit for sleep and wake up control
Pilot lampPower indicator: 1
Heat alarm light: 1
Wake up indicator: 1
Spare lights: 4
Compatible with Raspberry PI 2.0mm double row pin 2*20pinUART: 2 TTL3.3v levels (1 debug serial port)
CAN: 1 TTL3.3v level
I2C: 2 PCS
I2S: 1
SPI: 1
PWM: 1
GPIO: 9
Type-C Power portPower adapter requirements: Input: AC100-240V.50-60HZ, output: DC12V;
Type-C source, current 3A-5A; Recommended DC power supply or adapter above 12V/3A, ripple less than 120mv;
Operating ambient temperature-20℃ ~ 80℃（-4℉～176℉）
Typical power consumptionTypical power consumption of Atlas 200I A2 module: 25W / 21W;
Typical power consumption: 25W-30W;
Operating systemopenEuler 22.03、Ubuntu 22.04
Deep learning frameworkMindSpore
Compilation toolCCE/CCE compiler Tool



Previous: RockChip R3576-01 Developer Kit

Next: RockChip RK3588 Developer Kit

APPLICATION

Intelligent terminals are a type of embedded computer system equipment equipped with NVIDIAJetson series core modules, which have the characteristics of small size, high performance, low power consumption, and strong stability.