HiHope
Luna
The HiSilicon Boudica150 platform “Hi2115” is used to provide a good secondary development software and hardware environment, which can support dual-mode working model, the unique “NB-Iot stand alone” master control working mode, saves external MCU resources, and support the operation of light-weight cores such as the Huawei Lite OS, and the onboard supports a variety of NB-IoT antennas, which is convenient for users in different application environments.
Product Highlights
Localized design, built-in chip, supports multiple encryption algorithms
By using Hisilicon Boudica150 platform “Hi2115”, it provides a good software and hardware environment for secondary development
Supports dual-mode working mode as the unique “NB-Iot stand alone” master control working mode saves external MCU resources
Supports the operation of Huawei Lite OS operating system with a lightweight kernel
Onboard supports a variety of NB-IoT antennas, which is convenient for users in different applications
Rich applicable scenarios to meet the needs of users for various IoT applications
Applicable scenarios: Internet of Things, Smart Industry and Security System.
| Functional Configuration | |
| CPU | HiSilicon Hi2115 |
| SDRAM | 64KB |
| Flash | 352KB |
| Supported communication standard | China Mobile: Band 8; China Unicorm: Band 3/Band 8; China Telecom: Band 5; European/Southease Asia Operators: Band 28 |
| Secure encryption | Equipped with SM1, SM4, SM6, SM2, SM3 and other national cryptographic algorithms; also supports 3DES, AES, RSA 1024/RSA2048, SHA256 and other commonly used international cryptographic algorithms |
| Power supply | 5V/2A via MicroUSB / 2 pin battery connector |
| 40 pin expansion interface (pitch: 2.0mm) | 32 x GPIOs, UART, 12C, SPI, PWM |
| 30 pin expansion interface (pitch: 2.0mm) | 8 x GPIOs, UART, 12C, SPI, PWM, ADC |
| Antenna | On-board ceramic antenna/SMA jack |
| Other | JTAG (10pin) x 2; SIM card slot x 1; e-SIM; Micro USB x 2; Key; LED |
AKEBI96
New 4K/8K technologies such as HDR processing are used to realize a high-quality, high-performance and low-cost TV system. By adding an 8K decoder externally, it can also be expanded into an 8K TV system.
Product Highlights
Based on Socionext LD20 (SC1408AJ1) SoC
The processor adopts 2 ARM Cortex-A72 and 2X ARM Cortex-A53, equipped with 3GB LPDDR3-SDRAM memory and 16GB eMMC storage, supports Micro SD (SD3.0), PCI-E slot, 4K HD playback
The chip platform has a high degree of open source, and is the first recommended development platform for Google Android TV
Product Parameter
| Function Parameter | |
| SoC | SC1408AJ1 |
| CPU | ARM Cortex-A72x2 and Cortex-A53x2 |
| GPU | Mali-T820MP3 |
| RAM | 3GB DDR3 SDRAM |
| ROM | 16GB eMMC |
| OS | Android TV |
| Interfaces | Expansion slot |
| WiFi/BT | IEEE 802.11 a/b/g/n/ac with BLE4.2 |
HiKey970
Hikey970 is based on the world’s new mobile AI processing chip, HiSilicon Kirin 970 development board, supports Huawei HiAI artificial intelligence engine, and Google Android NN neural network computing framework.
As a partner of Huawei, HiKey 970, an artificial intelligence development platform jointly launched by HopeRun Software and Huawei, is an AI mobile computing platform based on built-in NPU. It integrates Huawei’s innovatively designed HiAI framework and other mainstream neural network frameworks. In addition to the support of CPU and GPU AI operations, it also supports NPU-based neural network computing hardware acceleration, enabling developers to implement deep learning algorithms, intelligent robots and the development of urban areas that would make smart life a reality.
HopeRun Software assists Huawei in promoting the process of open source and the promotion of the HiAI ecosystem, bringing advanced open source achievements to the HiKey 970 platform, attracting more developers to the HiSilicon platform, and relying on the HiKey 970 platform to form an end-to-end industry around AI solution.
HiKey 970 will provide developers with an open source development board with powerful AI computing power, such as hardware acceleration, enhancing performance, ecological construction of AI applications, and collaborations between developers to jointly create a bright future for mobile AI applications.
PRODUCT HIGHLIGHTS
Using Kirin970 SoC, the same processor as Mate10
Integrates Huawei’s innovative HiAI framework and other mainstream neural network frameworks. While supporting both CPU and GPU AI operations, it also supports NPU-based neural network computing hardware acceleration. The energy efficiency and performance can reach 50 times and 25 times that of CPU operations, respectively.
Provide complete multi-application mode and machine learning framework support, with more complete documentation, richer and more effective API, and with source code that can be used more effectively, allowing developers to understand the immense potential of AI.
Native support for Google AOSP and support for multiple Linux distributions such as Debian/Ubuntu
Applicable scenarios: Robot, Cars and Smart City
Parameter Details
| Function parameter | |
| Processor Series | Kirin |
| SoC | Hisilicon Kirin 970 |
| CPU | ARM Cortex-A73 x 4 and Cortex-A53 x 4 |
| GPU | ARM Mali-G72 MP12 |
| AI | Dedicated Neural-network Processing Unit (NPU) |
| RAM | 6GB LPDDR4 SDRAM |
| ROM | 64GB UFS flash storage |
| Micro SD card slot | |
| PCIE Gen2 on M.2 M Key connector | |
| Wireless | WiFi (2.4 -and 5-GHz dual band with two antennas) |
| Bluetooth 4.1 | |
| GPS, Glonass | |
| USB | 2 x USB 3.0 type A(host only) |
| 1 x USB 2.0 type C (OTG) | |
| 1 x USB 2.0 (host only) | |
| Display | 1 x HDMI 1.4 (type A - full) |
| 1 x 4L - MIPI DSI | |
| HDMI output up to FHD 1080P | |
| Video | MPEG2@80Mbps, MPEG4@50Mbps |
| H264@240Mbps, H265@160Mbps | |
| VP8@100Mbps, VP9@120Mbps | |
| Audio | HDMI output |
| Camera | 1 x 4-lane MIPI CSI |
| 1 x 2-lane MIPI CSI | |
| Expansion interface | 40 pin low speed expansion connector: |
| `+1.8V, +5V, DC power, GND, 2UART, 2I2C, SPI, I2S, 12*GPIO | |
| 60 pin high speed expansion connector: | |
| 4L-MIPI DSI, 12C x 2, SPI(48M), USB 2.0, 2L + 4LMIPI CSI | |
| CAN | Implements CAN V2.0B at 1Mb/s |
| Ethernet | Gigabit Ethernet |
| LED | 1 x WiFi activity LED |
| 1 x BT activity LED | |
| 4 x User LEDs | |
| Button | Power button: button power on / off & reset the system |
| Power source | Recommend a 12V@2A adaptor with a DC plug which has a 4.75mm outer diameter and 1.7mm center pin with standard center -positive (EIAJ-3 Compliant) |
| OS support | AOSP/LINUX |
| Appearance characteristic | 100mm x 85mm x 10mm |
HiKey960
HiKey960 is an AOSP development board officially certified by Google based on HiSilicon Kirin 960 chip. It supports new and existing sensors, cybersecurity technologies and their protocols, other peripheral hardware and software, and innovative capabilities.
Application scenarios: Digital Media, Intelligent Robots and Smart Payment
Product Highlights
Based on Kirin960 SoC, same processor as Mate9
Compliant with the Linaro 96Boards open development board specifications, and reserves a wealth of standard external I/O expansion interfaces, which can be applied to external I/O devices, displays and cameras in the 96Boards ecosystem
Not only supported by new and existing sensors, network security technologies and their protocols, other peripheral hardware/software along with innovative features, but also supports Google AOSP, multiple Linux distributions such as Debian/Ubuntu
Product Parameter
| Function Parameter | |
| SoC | Kirin 960 |
| CPU | ARM Cortex-A73 x 4 and Cortex-A53 x 4 |
| GPU | ARM Mali-G71 MP8 |
| RAM | 4GB LPDDR4 SDRAM |
| PMU | Hi6421GWCV530 PMU |
| ROM | 32GB UFS Flash Storage |
| Wireless | WiFi (2.4- and 5-GHz dual band with two antennas) and Bluetooth 4.1 |
| USB | 2 x USB 3.0 type A (host mode only) and 1 x USB 2.0 type C OTG |
| Display | 1 x HDMI 1.4 (Type A - full) |
| Video | Encoder:H.265/H.264 3840 x 2400@30fps 4x1080p @ 30 fps |
| Decoder: H.265, HEVC MP/High Tier, Main 10/High Tier, H.264 BP/MP/HP, MPEG1/2/4, VC-1, VP6/8 | |
| Audio | HDMI Output |
| Camera | 1 x 4-lane MIPI CSI and 1 x 2-lane MIPI CSI |
| Expansion interface | 40 pin low speed expansion connector: 4L-MIPI DSI, 12C x 2, SPI (48M), USB2.9, 2L+4LMIPI CSI |
| LED | 1 x WiFi activity LED (Yellow), 1 x BT activity LED (Blue) and 4 x users LEDs (Green) |
| Button | Power button: button power on/off & reset the system |
| Power source | Recommend a 12V@2A adapter with a DC plug which has a 4.75mm outer diameter and 1.7mm center pin with standard center-positive (EIAJ-3 Compliant) |
| OS | AOSP/LINUX |
| Size | 85mm x 55mm |
Poplar
Poplar is an Android TV development platform to showcase one’s creativity. The example of application is on Smart POS, vending machine and tablet.
Product Highlights
Based on Hisilicon Hi3798CV200 SoC
The processor adopts 4 ARM Cortex-A53 2.0 GHz, and is equipped with ARM Mali-T720 3D image processor, equipped with 2GB LPDDR3-SDRAM memory and 8GB eMMC storage, supports Micro SD (SD3.0), PCI-E plug-in slot, 4K HD playback
The new 96Boards TV development board will be natively supported by Android TV, and will also support Linux distribution boards such as Debian/Ubuntu
Parameter Details
| Functional Configuration | |
| SoC | HiSilicon Hi3798CV200 |
| CPU | ARM Cortex-A53 x 4 |
| RAM | 2GB DDR3 SDRAM |
| ROM | 8GB eMMC on-board |
| USB | One USB 2.0 port and one USB 3.0 port |
| Console | MicroUSB port for console support |
| Ethernet | 10M/100M/1000M Gigabit Ethernet |
| PCIe | One PCIe 2.0 interface |
| JTAG | 10-pin JTAG |
| Expansion interface | Linaro 96Boards 40 pin low speed expansion slot |
| Dimension | Standard 160 x 120mm 96Boards Enterprise Edition form factor |
| WiFi | 802.11AC 2x2 with Bluetooth |
| Connectors | One connector for Smart Card and one connector for TSI Card |
Musashi
The Musashi Microprocessor (MPU) features a 64-bit multi-core architecture with high graphics capabilities and a high bandwidth memory interface to support systems with enhanced human-machine interfaces. This can be applied to smart home and smart industry.
Product Highlights
Based on Renesas RZ/G2M SoC
The processor adopts 2X ARM Cortex-A57 and 4X ARM Cortex-A53, equipped with 4GB DDR4 memory and 32GB eMMC storage, supports Micro SD (SD3.0), PCI-E slot
The chip has powerful multimedia functions and image processing capabilities, and meets industrial-grade specifications
Parameter Details
| Function parameter | |
| Processor series | RZ/G Series |
| SoC | RZ/G2M |
| CPU | ARM Cortex-A57 x 2 and Cortex-A53 x 4 |
| GPU | Imagination Technologies PowerVR Series 6XT GX6250@700MHz |
| RAM | 4GB LPDDR4 SDRAM |
| ROM | 32GB eMMC |
| Wireless | WiFi (2.4 -and 5-GHz dual band with two antennas), Bluetooth 4.1 |
| USB | 1 x USB 3.0 type A (Host/Function) |
| 1 x USB 2.0 type A (Host), 1 x USB 2.0 | |
| Micro (Host/Function/OTG) | |
| Display | 1 x HDMI 1.4 (Type A - full), HDMI output up to FHD 1080P, 1 x 4L - MIPI DSI |
| Video | MPEG2@80Mbps, MPEG4@50Mbps |
| H264@240Mbps, H265@160Mbps | |
| VP8@100Mbps, VP9@120Mbps | |
| Audio | HDMI Output |
| Camera | 1 x 2-lane MIPI CSI, 1 x 4-lane MIPI CSI |
| Expansion interface | 40 pin low speed expansion connector: +1.8V, +5V, DC power, GND, 2UART, 2I2C, SPI, 12,12*GPIO |
| 60 pin high speed expansion connector: 4L-MIPI DSI, 12C x 2, SPI(48M), USB 2.0, 2L +4LMIPI CSI | |
| CAN | Implements CAN V2.0B at 1Mb/s: 0-8 byter length in the data field; standard and extended data and remote frames |
| Receive buffers, masks and filters: Two receive buffers with prioritized message storage; Six 29-bit filters; Two 29-bit masks | |
| Ethernet | Gigabit Ethernet |
| LED | 1 x WiFi activity LED, 1 x BT activity LED, 4 x User LEDs |
| Button | Power button: button power on/off & reset the system |
| Power source | Recommend a 12V@5A adaptor with a DC plug which has a 4.75mm outer diameter and 1.7mm center pin with standard center-positive(EIAJ-3 Compliant) |
| OS support | LINUX |
| Appearance characteristic | 100mm x 85mm |
Hickory
Hickory3559A adopts an embedded SoC chip (HiSilicon Hi3559AV100) and adopts 12nm ultra-low power consumption process. In addition to multi-core CPU and high-performance GPU, it also integrates dual-core NNIE neural network, with a computing power of up to 4T, setting the standard of Debian/Ubuntu system, providing good compatibility and efficient AI application building capabilities. This can be applied to smart home and smart industry.
Product Highlights
Provide digital video recording with 8K30/4K120 broadcast-level image quality, support multi-channel Sensor input/H.265 encoding output or film-level RAW data output, integrate high-performance ISP processing and adopt advanced low-power technology and low-power architecture designed to provide users with excellent image processing capabilities.
Supports a variety of neural networks (8 classification models, 7 target detection models, 3 long and short-term memory models), covering a wide range of AI application scenarios including classification, target detection, speech semantics, and behavior analysis.