Allwinner Technology partnerRadxa today launched the Cubie A7A single board computer based on the Allwinner A733 processor platform. This product combines high-performance computing, AI acceleration capabilities, and flexible interface expansion, targeting various application scenarios such as edge computing, artificial intelligence, industrial automation, and smart home, providing developers, enterprise customers, and system integrators with a cost-effective, long-lifecycle hardware foundation.
PowerfulPerformance
The Cubie A7A is equipped with the Allwinner A733 high-performance octa-core AI platform processor, which includes 2× Arm® Cortex®-A76 + 6× Arm Cortex-A55 CPU cores, complemented by an Imagination BXM-4-64 GPU, supporting graphics and computing standards such as OpenGL ES 3.2, Vulkan 1.3, and OpenCL 3.0. Additionally, it integrates a Vivante VIP9000 neural network processing unit (NPU), providing up to 3 Tops of AI inference computing power and supporting various precision operations such as INT8, INT16, FP16, and BF16, meeting diverse AI application needs. In terms of storage, the Cubie A7A supports up to 16GB LPDDR5 memory, as well as various high-speed storage interfaces including eMMC 5.1, UFS 3.0, and NVMe SSD, ensuring data processing speed and system responsiveness.
Rich Interfaces and Flexible Expansion
The Cubie A7A offers a rich set of peripheral interfaces including PCIe 3.0, HDMI 2.0, MIPI CSI/DSI, USB 3.1, Gigabit Ethernet (supporting PoE), and 40Pin GPIO, compatible with various peripheral and sensor modules. At the same time, it features built-in Wi-Fi 6 and Bluetooth 5.4, enabling diverse wired and wireless connectivity options for quick integration in different deployment environments.
Multi-System Support and Developer-Friendly Ecosystem
The Cubie A7A supports Debian Linux, Tina Linux, Buildroot, and Android 13 industry version operating systems, compatible with mainstream AI frameworks (including ONNX, TensorFlow, PyTorch, Caffe, TFLite, etc.), and provides the official Allwinner NPU development toolchain and supporting examples, significantly lowering the development and deployment threshold for AI applications. Additionally, the onboard RISC-V architecture XuanTie E902 coprocessor can run RTOS real-time tasks, achieving low-power standby, real-time control, secure boot, and other functions, further enhancing system stability and security.
Win-Win Cooperation to Promote Edge AI and Open Source Ecosystem Development
Radxa founder Tang Liang stated:
“Allwinner Technology has a deep accumulation in high-efficiency processor design and AI chip optimization, while Radxa has rich experience in single board computer design, hardware open ecosystem, and global developer community operation. The Cubie A7A is the result of the combination of both parties’ advantages, setting a new benchmark for edge computing. We hope to bring this high-performance edge AI platform to more developers and industry users through this product.”
Allwinner Technology’s Vice President of Solution OperationsZhu Zhenhua stated:
“The A733 is a high-performance platform launched by Allwinner for the new generation of edge computing and AI applications. We have long adhered to the strategic concept of embracing open source, open markets, and industry ecosystems. Our cooperation with Radxa in the field of single board computers is not only an innovation in product implementation but also an important step in promoting the healthy development of the open-source hardware ecosystem. In the future, Allwinner will continue to work with ecosystem partners to expand the application boundaries of the A733 in artificial intelligence, robotics, industrial control, and other niche fields.”
Market and Application Prospects
With its high-performance CPU+GPU+NPU combination, open hardware interfaces, mature software ecosystem, and long-term supply assurance (from 2025 to 2034), the Cubie A7A will have broad application space in fields such as robotic control, machine vision inspection, smart home gateways, and education and research, providing stable and sustainable hardware foundations for partners and developers.
About RadxaRadxa was established in 2013 and is a technology company focused on the research and development of single board computers, embedded computing platforms, and related ecosystem products, widely used in education, research and development, industrial control, and edge computing. Radxa adheres to open cooperation, actively participates in open-source community building, and is committed to providing high-performance, scalable, and long-term supply hardware platforms for global developers and enterprises.
Allwinner Technology “A Series” Product Trends
《Allwinner Launches Three Tablet Processors, Precisely Covering All Ranges of Android Tablets
《Allwinner A733 Processor Passes Google’s Latest Android 15 GMS Certification!
《Mobile Cloud Smart Education Tablet Powered by Allwinner A733Allwinner Chip News Collection
◾ A Series
<<<
◾ T Series
<<<
◾ MR Series
<<<
◾ V Series
<<<
◾ H Series
<<<
◾ F Series
<<<
◾ TV Series
<<<
◾ IoT Series
<<<
◾ Analog Series
<<<