APES Who Compute
Nvidia System on Module Carrier Board for Local AI Workloads
Nvidia System on Module Carrier Board for Local AI Workloads
Jesse Cheung and Nicholas Dalton
Overview:
This project is a carrier board for the Nvidia System on Module Orin Nano/NX series packed with 16GB of DDR5 RAM, 8-core Arm® Cortex®-A78AE CPU, 1024-core Ampere + 32 Tensor Cores GPU, and AI Performance 100 TOPS.
Description:
This project has peripherals such as M.2 Key M NVME SSD 8 Lane PCIE Gen 4, 2x M.2 Key E Wi-Fi and Bluetooth,1Gbit Ethernet RJ45, HDMI 2.1, USB Debug and USB 2.0 Controller, 40 Pin GPIO Header, and many sensors
3D Model of PCB prior to production
Technical Progress
Board Development
Board Stackup
6 Layer Board
Impedance Matched Stack-Up
Result
USB-UART Uploading
Power Delivery through Barrel Connector and USB-C
LED1 and LED2 controls