New AIBOX Member with RK3588S Chip Now Available

0 comments

The AIBOX product matrix has added a new member: AIBOX-3588S. This device supports the private deployment of large AI models and has 32GB LPDDR5 ultra-large memory. It is widely used in edge computing, private deployment of large models, smart security, smart home, smart industry, and other fields.

Octa-core processor RK3588S2

RK3588S is Rockchip's flagship AIoT chip, which adopts an 8nm LP process. It has an octa-core 64-bit CPU with a main frequency of up to 2.4GHz. It integrates an ARM Mali-G610 MP4 quad-core GPU and a built-in AI accelerator NPU, bringing more optimized performance for various AI application scenarios.

RK3588S2

Private deployment of large AI models

The NPU computing power can reach 6 TOPS, supporting the private deployment of modern mainstream large models, such as the Gemma series, ChatGLM series, Qwen series, Phi series and other large language models; supporting the import and export of RKNN models, and also supporting TensorFlow, TensorFlow Lite, PyTorch, Caffe and other deep learning frameworks.

large AI models

32GB LPDDR5 large memory

Compared to LPDDR4, LPDDR5 delivers greater memory capacity, higher bandwidth, faster data transfer rates, lower power consumption, and more advanced Error Correction Code (ECC) technology. This makes it ideal for the private deployment of large AI models, meeting the demands for both memory space and response speed. By maximizing hardware synergy, LPDDR5 enables models to run more efficiently, enhancing overall system performance and energy efficiency.

32GB LPDDR5 large memory

Supports 8K video decoding

Supports 8K@60fps H.265/VP9 video decoding and 8K@30fps H.265/H.264 video encoding, with simultaneous encode-decode capability. It can handle up to 32 channels of 1080P@30fps decoding and 16 channels of 1080P@30fps encoding. The high-resolution and multi-channel decoding power enables large AI models to rapidly extract video data, providing richer inputs for model training and inference. This enhances visual analysis accuracy while accelerating algorithm training and optimization.

Supports 8K video decoding

Full metal casing, efficient heat dissipation

Configured with an industrial-grade all-metal casing, the aluminum alloy structure enhances heat conduction. The top cover and side panels feature a ribbed grille design to ensure external airflow and efficient heat dissipation, maintaining computing performance and stability even under high-temperature operation. Compact and refined, the machine operates stably to meet various industrial-grade application requirements.

Full metal casing, efficient heat dissipation

Complete development materials

Support Linux OS, providing a safe and stable system environment for product research and development; providing supporting source code, tutorials, technical information and development tools to make development easier and more convenient.

Wide range of application scenarios

It is widely used in scenarios such as intelligent monitoring, AI teaching, computing services, edge computing, private deployment of large models, data security and privacy protection, etc.


Firefly debuts at Chengdu International Industry Fair 2025

Leave a comment

Please note, comments need to be approved before they are published.