The Ultimate AI Duo: RK1820 & RK3588
Understanding How the RK1820 Co-Processor Collaborates with the RK3588 Why is there increasing talk about NPUs and co-processors in edge devices? While the RK3588 is a powerful SoC with
Neardi single board computer include three series, namely Rockchip Series, Nvidia Series, and NXP Series, which are developed with Rockchip, Nvidia, NXP, and other brand chips as the core CPU.
Home » LKD3588S Single Board Computer




The LKD3588S is a multifunctional development evaluation board carefully designed based on the Rockchip RK3588 chip platform. It consists of our LCB3588S system on modules and baseboards. The core module and baseboard are connected by B2B connectors and secured with four M2 screws, ensuring stability and reliability.
This compact and slim board offers various functions and rich interfaces, making it suitable for products with limited structural space.
The LKD3588S features 3 Type-A USB 3.0 HOST ports, 1 Type-C USB 3.1 OTG port, and 1 4-pin PH2.0 USB 2.0 HOST interface onboard, which can connect multiple USB cameras. It also includes 2 mini-PCIe interfaces that support external 4G modules and RK1808-based mini-PCIe interface NPU computing cards. Additionally, the LKD3588S supports common communication module interfaces such as dual-band WiFi 6, BT 5.0, dual-channel Gigabit Ethernet, UART, I2C, RS232, RS485, and CANBUS.
For display options, the board supports 3 HDMI outputs, 1 dual-channel LVDS output, 1 DP interface output, and multi-screen display capabilities. It can also accommodate multiple MIPI-CSI camera interface inputs and 1 HDMI 2.0 interface input.
Function | Description |
CPU | RK3588S, quad-core Cortex-A76@2.4GHz + quad-core Cortex-A55@1.8GHz |
GPU | ARM Mali-G610 MC4;OpenGL ES 1.1/2.0/3.1/3.2;Vulkan 1.1/1.2 OpenCL 1.1/1.23/2.0 |
NPU | 6 TOPS, supports int4/int8/int16/FP16/BF16/TF32 |
VPU | Decode: H.265/H.264/AV1/VP9/AVS2@8K60FPS Encode: H.264/H.26522@8K30FPS |
DDR | 4GB/8GB/16GB LPDDR4 |
eMMC | eMMC 5.1, 32GB/64GB/128GB(Optional) |
System | Android / Ubuntu / Buildroot / Debian |
Size | 160*115*28.25 mm |
Hardware Interface | |
Power | DC12V – 3A (DC Jack 5.5*2.1mm / PH2.0 wafer connector) |
USB | 3*Type-A USB3.0 HOST 1* Type-C USB3.1 OTG 1*4Pin PH2.0 USB2.0 HOST |
Display | 2*Type-A HDMI 2.1 up to 8K@60fps or 4K@120fps 1*Type-A HDMI 1.4 up to 1080P@60fps Duel channel LVDS up to 1080P@60HZ 1*HDMI-IN(4K@60fps),支持 HDCP 2.3 |
Audio | φ3.5mm earphone Jack with L/R audio out φ3.5mm micphone Jack with Mic in 1*HDMI audio out |
Camera | 2* MIPI CSI (4 Lane) or 4*MIPI CSI (2 Lane) + 2* MIPI CSI (4 Lane) |
Mini-PCIe | mini PCIe for 2G/3G/4G/5G module RK1808 AI computing card |
SD card | Compatible with SDIO 3.0 protocol, system boot up supported |
SIM card | Micro sim slot for Mini-PCIe 4G LTE module |
RJ-45 | 2*10/100/1000-Mbps data transfer rates |
RTC | RTC power on and off supported |
Serial port | 3*Uart, 1*I2C |
Keys | 3* keys (power, reset, update) |
Power output | 12V, 5V, 3.3V,1.8V |
The B2B connection and secure mounting with M2 screws enhance stability and durability, making the LKD3588S ideal for applications in high-vibration environments or with space limitations.
The LKD3588S’s mini-PCIe slots support 4G modules and NPU computing cards, such as the RK1808, expanding connectivity options and AI processing power for edge computing.
It supports low-power modes, programmable shutdown, and efficient cooling to prevent overheating, essential for continuous or 24/7 industrial deployments.
Due to its slim design, using fans or heatsinks on high-load areas like the RK3588 processor or mini-PCIe slots is recommended. Heat management is key for high-demand applications.
Understanding How the RK1820 Co-Processor Collaborates with the RK3588 Why is there increasing talk about NPUs and co-processors in edge devices? While the RK3588 is a powerful SoC with
Hi3516DV500/Hi3519DV500 Product Introduction The integration of AI into security is accelerating: AI vision has now fully permeated fields such as intelligent surveillance, edge computing, access control terminals, vehicle recognition, and
As edge AI applications become increasingly widespread, a core challenge for developers is enabling AI models to run quickly and efficiently on edge devices. Addressing complex AI algorithms and computational