Core-3588JD4

AI Large-Model Core Board

Octa-core AI Processor RK3588

Core-3588JD4 is equipped with the Rockchip flagship processor RK3588, an octa-core 64-bit CPU capable of
delivering 6 TOPS of computing power, enhancing performance for AI application scenarios.

Equipped with the aBMC management system

upports the private deployment of expansive parameter models based on the Transformer framework,
including substantial language models like Gemma-2B, ChatGLM3-6B, Qwen1.5-1.8B, and Phi-3-3.8B.

Decoding of 8K
Ultra-High Definition Video

Enables 8K high-definition video decoding, delivering clearer images
and richer details, with a 2D hardware engine that significantly
enhances display performance.

32 GB of Massive Memory

Surpassing previous memory capacity limitations, it offers faster
response speeds and meets the demands of applications that
require substantial memory and large storage capacities.

Industrial-Grade Stability

Supports external watchdog for industrial-grade stability. Compatible
with multiple operating systems, suitable for ARM PCs, edge computing,
cloud servers, intelligent NVRs, and other fields.

Comprehensive Expansion Interfaces

Specifications

Core-3588JD4
Basic Specifications
SOC

Rockchip RK3588

CPU

Octa-core 64-bit processor (4×Cortex-A76+4×Cortex-A55) , main frequency up to 2.4GHz

GPU

ARM Mali-G610 MP4 quad-core GPU, support OpenGL ES3.2/OpenCL 2.2/Vulkan1.1, 450 GFLOPS

NPU

The computing power is up to 6TOPS(INT8), support INT4/INT8/INT16 mixed operations

ISP

Integrated 48MP ISP, support HDR and 3DNR

Codecs

Decoding: 8K@60fps H.265/VP9/AVS2, 8K@30fps H.264 AVC/MVC, 4K@60fps AV1, 1080P@60fps MPEG-2/-1/VC-1/VP8 Encoding: 8K@30fps H.265/H.264

RAM

LPDDR4/LPDDR4x (4GB/8GB/16GB optional, up to 32GB)

Storage

eMMC (32GB/64GB/128GB/256GB optional)

Power

5V (voltage tolerance ± 5%)

Power consumption

Max: 13W(5V/2600mA), Normal: 2.8W(5V/560mA), Min: 0.175W(5V/35mA)

OS

Android, Linux OS

Software Support

· Support the privatization deployment of ultra-large-scale parametric models under the Transformer architecture, such as Gemma-2B, ChatGLM3-6B, Qwen-1.8B, Phi-3-3.8B and other large language models · It supports traditional network architectures such as CNN, RNN, and LSTM, and supports the import and export of RKNN models; Support a variety of deep learning frameworks, including TensorFlow, TensorFlow Lite, PyTorch, Caffe, ONNX and Darknet. It also supports the development of custom operators · Support Docker container management technology

Interface

SODIMM (260 PIN, 0.5mm pitch)

Size

69.6mm × 45.0mm × 4.6mm

Interface Specifications
Weight

≈18g

Environment

Operating Temperature: -20℃~60℃, Storage Temperature: -20℃~70℃, Storage Humidity: 10%~90%RH(non-condensing)

Internet

2 × Gigabit Ethernet (MDI interface is provided, and the core board has an onboard Ethernet PHY chip) Expandable WiFi6/Bluetooth 5.2 via SDIO3.0/PCIe3.0 Expandable 5G/4G LTE via USB3.1 (Gen1)/USB2.0

Video input

MIPI CSI (2×4Lanes/4×2Lanes/1×4Lanes + 2×2Lanes)

Video output

1 × HDMI2.1 TX/eDP1.3 TX (8K@60Hz, HDMI supports HDCP2.3; Supports eDP1.3, 4K@60Hz, supports HDCP1.3; HDMI and eDP cannot work at the same time)

Audio output

2 × I2S (2 channels), 2 × SPDIF, 1 × PDM (8 channels, support multi-MIC array)

USB

2 × USB3.1(Gen1)OTG、1 × USB3.1(Gen1)HOST、2 × USB2.0 HOST、2 × USB2.0 OTG

PCIe

1 × PCIe3.0 (2×2lanes, 1×4lanes, 4×1lanes) 、3 × PCIe2.1 (1 lane)

SATA

3 × SATA3.0 (Multiplexed with PCIe 2.1)

Watchdog

Independent watchdog

Other interfaces

8 × I2C、7 × UART、4 × SPI、2 × ADC、15 × PWM、1 × SDMMC、2 × CAN、GPIO