RD-N1-Edge Dual-Chip platform software user guide
Introduction
RD (Reference Design) is a collection of resources to provide a representative view of typical compute subsystems that can be designed and implemented using specific generations of Arm IP. RD-N1-Edge in particular is based on the following hardware configuration.
8x Neoverse N1 Cores with DynamIQ Shared Unit (DSU)
Dedicated L2 cache: 512KB per core
Shared L3 cache: 2MB per cluster
CMN-600 with CML option: 8MB System Level Cache and 16MB Snoop Filter
DMC-620 with 2xRDIMM DDR4-3200
Multiple AXI expansion ports for I/O Coherent PCIe, Ethernet, offload
Arm Cortex-M7 for System Control Processor (SCP) and Manageability Control Processor (MCP)
RD-N1-Edge dual-chip is a platform configuration in which two RD-N1-Edge platforms are connected through high speed CCIX link. The CCIX link is enabled by CMN600’s Coherent Multichip Link. Such platforms are called RD-N1-Edge-Dual hereafter.
This document is a user guide on how to setup, build and run the RD-N1-Edge-Dual software stack on the Fixed Virtual Platform. The components integrated into this stack is described in the software stack introduction page.
Setting up the Host Machine and downloading the software stack
Host Machine requirements and the necessary steps to be followed to sync the software stack are listed on ‘Setup Workspace’ page.
Obtaining the RD-N1-Edge-Dual FVP
The latest version of the RD-N1-Edge-Dual fixed virtual platform (FVP) can be downloaded from the Arm Ecosystem FVPs page. On this page, navigate to “Neoverse Infrastructure FVPs” section to download the RD-N1-Edge platform FVP installer.
Follow the instructions of the installer and setup the FVP. The installer, by default, selects the home directory to install the FVP. To opt for different directory than the one selected by the installer, provide an absolute path to that directory when prompted for during the FVP installation process.
Supported features
RD-N1-Edge-Dual platform software stack supports the following features.
Follow the links above for detailed information about the build and execute steps for each of the supported features.