Home MYIR | Single Board Computers FZ5 Card – AI Accelerator Card
MYIR | Single Board Computers

FZ5 Card – AI Accelerator Card

Xilinx Zynq UltraScale+ ZU5EV MPSoC AI Accelerator Card with up to 2.4TOPS performance, 8GB DDR4, and 32GB eMMC storage. Supports PetaLinux and Baidu's PaddlePaddle AI Framework..

  • Detail
  • Special Features
  • Specification

The FZ5 Card is an AI accelerator card based on Xilinx’s Zynq UltraScale+ ZU5EV MPSoC. It offers high-performance computing with 1.5 GHz quad-core ARM Cortex-A53 and 600 MHz dual-core ARM Cortex-R5 processors. Integrated with ARM Mali-400 MP2 GPU and H.264/H.265 Video Codec Unit (VCU), it supports up to 2.4TOPS AI performance and runs at 55 FPS for ResNet-50.

The card comes with 8GB DDR4 memory, 32GB eMMC, and 64MB QSPI Flash, allowing for efficient handling of AI tasks. Connectivity options include RS232, RS485, USB 3.0, Gigabit Ethernet, CAN, and HDMI-IN for display and video input.

It is optimized for edge applications such as intelligent security, industrial testing, medical diagnosis, UAV inspections, and scientific research. The FZ5 Card supports running PetaLinux and Baidu’s PaddlePaddle AI framework, enabling easy AI model deployment.

  • Processor: 1.5 GHz Quad-core ARM Cortex-A53 and 600 MHz Dual-core ARM Cortex-R5 cores.

  • Graphics & Video: ARM Mali-400 MP2 GPU, H.264/H.265 Video Codec Unit (VCU).

  • Memory & Storage: 8GB DDR4 SDRAM, 32GB eMMC Flash, 64MB QSPI Flash, 32KB EEPROM.

  • Connectivity: RS232, RS485, 4 x USB 3.0, Gigabit Ethernet, CAN, TF, DisplayPort (DP), HDMI-IN, USB-UART, JTAG.

  • AI Performance: Up to 2.4TOPS, 55 FPS for ResNet-50.

  • Video Decoding & Analysis: Supports 8- to 16-channel Video Decoding and 4- to 8-channel Intelligent Analysis.

  • Software Support: PetaLinux, Baidu’s PaddlePaddle AI Framework.

  • Processor: 1.5 GHz Quad-core ARM Cortex-A53, 600 MHz Dual-core ARM Cortex-R5

  • Graphics: ARM Mali-400 MP2 GPU

  • Video Processing: H.264/H.265 Video Codec Unit (VCU)

  • Memory: 8GB DDR4 SDRAM (64-bit, 2400 MHz)

  • Storage: 32GB eMMC, 64MB QSPI Flash, 32KB EEPROM

  • I/O Ports: RS232, RS485, CAN, USB 3.0 (4x), Gigabit Ethernet, TF, DisplayPort (DP), HDMI-IN, USB-UART, JTAG

  • User I/O: 5 LEDs (Power, Status), 2 Buttons (Reset, User)

  • Power Supply: DC 12V/3A

  • Operating Temperature: -40°C to +85°C

  • Dimensions: 107mm x 96mm

  • Software Support: PetaLinux, Baidu PaddlePaddle AI Framework