How NoC architecture solves MCU design challenges
By Andy Nightingale, Arteris
EDN | April 16, 2025
Microcontrollers (MCUs) have undergone a remarkable transformation, evolving from basic controllers into specialized processing units capable of handling increasingly complex tasks. Once confined to simple command execution, they now support diverse functions that require rapid decision-making, heightened security, and low-power operation.
Their role has expanded across industries, from managing complex control systems in industrial automation to supporting safety-critical vehicle applications and power-efficient operations in connected devices.
As MCUs take on greater workloads, the conventional bus-based interconnects that once sufficed now limit performance and scalability. Adding artificial intelligence (AI) accelerators, machine learning technology, reconfigurable logic, and secure processing elements demands a more advanced on-chip communication infrastructure.
To meet these needs, designers are adopting network-on-chip (NoC) architectures, which provide a structured approach to data movement, alleviating congestion and optimizing power efficiency. Compared to traditional crossbar-based interconnects, NoCs reduce routing congestion through packetization and serialization, enabling more efficient data flow while reducing wire count.
To read the full article, click here
Related Semiconductor IP
- Smart Network-on-Chip (NoC) IP
- FlexNoC 5 Interconnect IP
- FlexNoC Functional Safety (FuSa) Option helps meet up to ISO 26262 ASIL B and D requirements against random hardware faults.
- Network-on-Chip (NoC)
- NoC Verification IP
Related Articles
- A Multi-Objective Optimization Model for Energy and Performance Aware Synthesis of NoC Architecture
- Design Challenges Drive Need for New Routing Architecture
- An HDTV SoC Based on a Mixed Circuit-Switched / NoC Interconnect Architecture (STBus/VSTNoC)
- Selecting an embedded MCU: How to avoid evaluation trap?
Latest Articles
- Making Strong Error-Correcting Codes Work Effectively for HBM in AI Inference
- Sensitivity-Aware Mixed-Precision Quantization for ReRAM-based Computing-in-Memory
- ElfCore: A 28nm Neural Processor Enabling Dynamic Structured Sparse Training and Online Self-Supervised Learning with Activity-Dependent Weight Update
- A 14ns-Latency 9Gb/s 0.44mm² 62pJ/b Short-Blocklength LDPC Decoder ASIC in 22FDX
- Pipeline Stage Resolved Timing Characterization of FPGA and ASIC Implementations of a RISC V Processor