DFA - Data Flow Architecture IP

Our distribution partners

DMU - DMA Unit | Supports DMA transfers between M_CAN Message RAM and System Memory

High-performance, deterministic hardware IP for real-time signal processing, control algorithms, and embedded AI acceleration.

DFA is a configurable hardware IP that accelerates real-time signal processing, control algorithms and embedded AI workloads. Its scalable architecture enables efficient integration across automotive MCU and SoC platforms while supporting deterministic, low latency execution.

By offloading compute intensive workloads from the CPU, DFA improves system efficiency, reduces latency and power consumption, and enables flexible, scalable compute solutions for next-generation automotive platforms.

Benefits

Deterministic real-time execution

CPU offloading for compute-intensive workloads

Low-latency, energy-efficient processing

Scalable architecture across performance points

Safety-ready integration for automotive platforms

Flexible acceleration for signal processing, control, and AI workloads

Technical Features

Product Type	Configurable Hardware IP
Product Type Integration	Configurable Hardware IP Soft IP (RTL)
Product Type Safety	Configurable Hardware IP ASIL-D capable
Product Type Target Platforms	Configurable Hardware IP MCU & SoC
Product Type Compute Engine	Configurable Hardware IP Floating-point Data Flow Architecture
Product Type Application	Configurable Hardware IP Motor Control, Real-Time Signal Processing, Sensor Fusion, Embedded AI

Downloads

DFA CDD Driver

GitHub Repository

Download

IP Techdays

IP Techdays provide a perfect exchange platform for OEMs, TIER 1, semiconductor vendors, eco environment providers, and other interested parties worldwide.

Attendees can share information and actively participate in discussions covering the latest developments in the IP environment.

IP Techday Asia 2026

September 16 + 17, 2026 in Shanghai, China

> click here for event details and presentations held

IP Techday Asia 2024

March 14 + 15, 2024 in Shanghai, China