Recently, I reviewed the design materials for the Allwinner A733 and found that the performance of this SoC is quite impressive. The hardware design-related materials have been uploaded to a cloud drive for everyone to study and research, including schematics, PCB files, IBIS simulation models, and design guidelines.
The Allwinner A733 is a high-performance octa-core AI platform processor aimed at mid-to-high-end tablets, interactive display devices, and various edge computing applications. This chip excels in performance, energy efficiency, and intelligent applications, thanks to its advanced 12nm process technology, powerful CPU and GPU configurations, and particularly its integrated 3 TOPS AI computing NPU.
Core Specifications and Performance
The Allwinner A733 achieves a balance between performance and power consumption in its architectural design, with the following main specifications:
CPU Architecture: It features an octa-core CPU with a big.LITTLE design, consisting of two high-performance Arm Cortex-A76 cores and six energy-efficient Arm Cortex-A55 cores, with a maximum clock speed of up to 2.0GHz. Additionally, it integrates a RISC-V architecture Xuantie E902 core for handling real-time tasks and low-power control.
GPU: It is equipped with the Imagination PowerVR BXM-4-64 MC1 GPU, supporting graphics and compute standards such as OpenGL ES 3.2, Vulkan 1.3, and OpenCL 3.0, providing smooth support for high-definition displays and gaming.
AI Computing Power: The built-in NPU has a computing power of up to 3 TOPS and supports various computation precisions such as INT8, INT16, FP16, and BF16. This enables the A733 to efficiently run various AI algorithms, such as AI Super Resolution (AI-SR, enhancing video clarity to 4K), AI ISP (image enhancement, improving low-light noise reduction by two levels), and gesture control, and even deploy and run large language models like DeepSeek-R1 1.5B on the device.
Memory and Storage: It supports LPDDR4, LPDDR4x, and the latest LPDDR5 high-speed memory, with a maximum capacity of 16GB. It also supports UFS 3.0 and eMMC 5.1 high-speed flash interfaces, significantly improving system read/write speeds and application launch smoothness.
Audio and Video Capabilities: It has powerful video encoding and decoding capabilities, supporting video decoding formats up to 8K@24fps for H.265/VP9/AVS2, and video encoding at 4K@30fps for H.265/H.264. In terms of display output, it supports dual-screen display and offers various interfaces such as RGB, LVDS, MIPI-DSI, eDP1.4, and DP, with a maximum support for 4K resolution display.
In terms of performance benchmarks, tablets equipped with the Allwinner A733 scored over 320,000 points in AnTuTu tests. Some evaluations suggest that its performance level is comparable to Qualcomm Snapdragon 778G or on par with MediaTek G88/Snapdragon 680, sufficient to meet the needs of smooth daily usage.
Main Application Areas
With its comprehensive performance configuration and strong AI capabilities, the Allwinner A733 has been widely used in various smart terminal devices.
Tablets: This is the core application market for the A733. Several tablet models have adopted this processor, such as the Whitedeer tablet and Acer’s Iconia A14/A16, targeting mid-to-high-end audio-visual entertainment and light office experiences.
Single Board Computers and Edge Computing: Targeting edge computing, industrial automation, smart home gateways, and other scenarios.
Interactive Displays: The A733 supports dual-screen display and various display interfaces, making it an ideal choice for interactive advertising machines, smart displays, and other applications.
In the future, Allwinner Technology and its ecosystem partners will continue to promote the application of the A733 in fields such as artificial intelligence, robotics, machine vision, and industrial control.
About the A733: Specific technical parameters, including the Datasheet, hardware design guidelines, stacked impedance design requirements, and reference documents such as schematics and PCB reference designs, as well as simulation IBIS models, are included in the materials shared by me. Interested individuals can refer to the resource acquisition method provided at the end to message Wu Chuanbin on his blog’s WeChat public account.
The reference design shared here only includes hardware design-related content and does not include software development-related SDKs or operating system image files.
Disclaimer
The document shared here comes from a user’s contribution and is not directly obtained from the original manufacturer. I have not verified it through actual mass production, and I do not guarantee the accuracy of the materials. Therefore, this information is for personal learning of PCB layout design purposes only and should not be used in a production environment.
How to Download Allwinner A733 Hardware Reference Design Materials
Follow the public account of my blog, and send the correspondingdownload keyword in the public account to obtain the download link.
Follow Wu Chuanbin’s blog public account
Message me on the public account:
Download|Allwinner A733 Materials
or the code
6396
It is recommended to copy and paste to avoid typos, O(∩_∩)O~