Embedded systems interface showing ARM Cortex model configuration, memory mapping, and data visualization tools

Overview

Features and Benefits

AI Debug Assistant for Multicore

Debug smarter with an AI assistant that connects directly to your live debug sessions via the Model Context Protocol (MCP). Compatible with GitHub Copilot, Claude Code, and any MCP-supported AI client, the debug assistant provides intelligent, context-aware debugging capabilities:

  • Natural-language hardware inspection, fault diagnosis, and root cause analysis
  • Autonomous investigation of crashes, memory corruption, and peripheral misconfiguration through 10 pre-built diagnostic prompts
  • Inspect registers, memory, variables, and stack traces with dedicated ARM Cortex-M and RISC-V fault decoding
  • Set breakpoints, conditional breakpoints, and hardware watchpoints via natural language commands
  • Execute raw GDB commands with built-in safety guards against destructive operations
  • Search workspace source code and analyze embedded C/C++ for common bug patterns
  • View RTOS threads and cross-core interactions in a unified environment
  • Support for both live debugging and post-mortem core dump analysis

Embedded AI Workflow Integration

Streamlined AI model deployment from training to production.

  • Import models via GUI or CLI for rapid integration
  • Automatic compatibility verification with ADI processors and microcontrollers
  • Profile runtime performance using Zephyr-based tools for latency and power analysis
  • Generate optimized, inference-ready code directly within the IDE
  • Deploy across low-power MCUs and high-performance DSPs

System-Level Orchestration

Simplify complex system configuration with visual resource management:

  • Allocate memory and peripherals visually across multiple cores and devices
  • Use JSON-based configuration for reproducible and version-controlled workflows
  • Central dashboard simplifies system-wide resource planning and debugging
  • Supports heterogeneous SoCs with ARM® TrustZone® partitioning
  • Reduces manual setup errors and accelerates bring-up
  • Graphical resource allocation for pins, clocks, power modes, and middleware

How It Works

Embedded developers face constant challenges with fragmented toolchains, manual configuration, and time-consuming debugging. CodeFusion Studio addresses these pain points with a unified development environment built on Visual Studio Code and deeply integrated with ADI's hardware ecosystem. From initial system planning through multi-core orchestration, Zephyr RTOS integration, and end-to-end AI deployment, CodeFusion Studio provides the tools you need to move faster, debug smarter, and build with confidence.

With support for multi-core orchestration, Zephyr RTOS, and AI end-to-end workflows, CodeFusion Studio helps you move faster, debug smarter, and build confidently.

AI Debug Assistant (Preview)

Debug smarter with an AI assistant that connects directly to your live debug sessions. Set breakpoints, inspect registers and memory, analyze hard faults, and investigate multi-core issues using natural language — powered by the Model Context Protocol (MCP) and compatible with GitHub Copilot and Claude Code.

IDE debugging C code, highlighting a buffer overflow crash analysis.


Embedded AI Tools

Manage AI models from import to deployment with built-in compatibility, profiling, and code generation tools.

AI model management interface for Arm Cortex-M4 and CNN Accelerator in embedded development environment


System Planner

Visually orchestrate pins, clocks, peripherals, memory, and inter-core data flows across supported SoCs all from one central dashboard.

Embedded development interface displaying memory block allocation and partition configuration for multiple cores


Workspace Creation Wizard

Start projects faster with a guided setup that integrates SoC selection, templates, and multi-core configuration in one flow - reducing setup time and ensuring consistency across teams.

Core project selection screen in CodeFusion Studio for Arm Cortex M4 and RISC-V setup

ADI CodeFusion Studio

Person working at a dual-monitor setup displaying data management and software interface dashboards Step 1:
Install Visual Studio Code, version 1.100 or later.

Download Visual Studio Code

Step 2:
Launch VS Code, go to Extensions, and install the CodeFusion Studio extension.

Install CodeFusion Studio

Step 3:
Download and install CodeFusion Studio tools and MSDK for macOS, Windows, and Linux.



Downloads & Related Software

Software Development Kit

AutoML for Embedded

AutoML for Embedded, part of the CodeFusion Studio™ ecosystem, empowers developers to train and deploy efficient AI models on resource-constrained platforms.

Get Started: Visit ADI Developer Portal

Visit ADI Developer Portal to begin your CodeFusion Studio download and accelerate your embedded development today.

Systems Requirements

CodeFusion Studio extension is supported on the following host operating systems:

  • Windows 11 (64-bit)
  • macOS 15 macOS 26 (ARM64)
  • Ubuntu 22.04 and 24.04 (64-bit)

Related Hardware (2)

Controller Boards

EV-Kits