Industrial and IoT products face constant pressure to become smaller, faster, and more connected. A decade ago, many industrial designs commonly paired an STM32 microcontroller with an Anybus module to provide application control and industrial communications. The approach allowed platforms to support multiple protocols by swapping modules. The architecture was reliable and flexible, but today’s requirements for performance, footprint, and long-term availability are driving engineering teams to adopt next-generation system-on-chips (SoCs) such as the Hilscher netX90.
What makes migration compelling is not only the need to avoid module obsolescence, but the opportunity to integrate application and communication processing in one device, reduce board size, and simplify compliance with industrial networking standards.
At DE Design Works, we’ve guided clients through this exact transition. From motion control to valve systems to industrial IoT devices, our team has firsthand experience mapping legacy architectures to modern platforms. Here’s what we’ve learned about why this migration matters and what technical details teams should pay attention to along the way.
Key Takeaways
- Industrial IoT designs increasingly require a single SoC that integrates application processing and industrial protocol communication.
- The netX90 eliminates the need for a separate Anybus module by integrating both an application core and a communication core on one chip.
- Pre-certified communication stacks (PROFINET, EtherNet/IP, EtherCAT) in the netX90 reduce development time and third-party certification costs.
- A single-chip architecture simplifies firmware toolchains, variant management, and over-the-air update strategies.
- Migrating from dual-chip to SoC-based designs reduces PCB footprint, power consumption, and long-term BOM cost.
From Two Chips to One

Traditional architectures split responsibilities: an STM32 handled the application, while an Anybus module managed industrial communications. It worked, but came with added complexity:
• Firmware had to manage an external SPI or parallel bus to the Anybus module.
• Two separate toolchains and update strategies were required.
• Memory and timing constraints limited protocol performance.
• Hosting both the STM32 and the Anybus module increased component count and PCB size requirements.
With SoCs like the netX90, the application core and communication core are integrated in a single device. The communication core runs pre-certified stacks for EtherCAT, PROFINET, EtherNet/IP, and others, while the application core handles product logic. This integration lowers BOM cost, simplifies firmware maintenance, and reduces risk during compliance testing. It also saves months of engineering and compliance work, since each new product variant does not require full protocol stack validation. By eliminating the external bus between MCU and module, latency is reduced and determinism improves in real-time applications such as motion control. Finally, having fewer components and interconnects increases reliability in high-vibration or temperature-sensitive environments.
STM32 to netX90 Migration Benefits
Business and technical pressures can make migration essential. Component end-of-life announcements, rising costs, and compliance requirements can disrupt product plans. Integrated SoCs like the netX90 also bring clear engineering advantages:
- Simpler design: One chip combines application and communication cores.
- Built-in protocol stacks: EtherCAT, PROFINET, and EtherNet/IP are ready without extra certification.
- Flexibility: Switch protocols via firmware licensing rather than redesign.
- Efficiency: Lower power, smaller size, and greater reliability with fewer interconnects.
While both approaches can support multiple protocols, they differ in how maintenance and product variants are handled. With Anybus modules, changing protocols often involves managing different hardware SKUs and running product-level requalification, even if the modules share a common pinout. With the netX90, protocol changes are handled through license-based firmware updates on a single hardware platform. In both cases, OEMs still need to validate their own products, but the scope of work can be narrower with an integrated SoC.
For OEMs, these advantages translate to faster time-to-market and the ability to serve diverse customers without separate hardware SKUs. They also reduce long-term maintenance, as engineers can focus on application features rather than supporting aging communication modules.
Key Considerations in the Transition
Migration requires structured engineering. A few of the most critical areas include:
- Firmware re-architecture: Application tasks that run on the STM32 must be ported to the APP core of the netX90. Interfaces such as ADIs (Application Data Interfaces) must be mapped cleanly to the COM core.
- Memory mapping: The STM32F4 family and the netX90 organize Flash and SRAM differently between application and communication tasks. Careful analysis is required to ensure code and buffers are sized and allocated correctly.
- Peripheral migration: GPIOs, timers, ADCs, and serial buses often need reassignment or reinitialization to match the new SoC’s capabilities.
- Toolchain differences: On STM32/Anybus designs, engineers typically used IDEs such as IAR, Keil, or STM32CubeIDE along with ST‑LINK utilities for the MCU, plus HMS configuration and firmware update tools for the module. By contrast, migration to netX90 means using Hilscher’s Eclipse‑based NetX Studio CDT and unified debug/programming tools for both cores. This reduces the number of separate toolchains to maintain, even though product-level validation is still required.
- Testing & validation: Regression testing against known protocol behaviors is essential. Packet timing, jitter, and failover behavior must be confirmed.
- Field updates: The netX90 bootloader supports firmware updates to both cores. Migrating teams must design an update process suitable for both production line programming and field service.
Certification Challenges
Industrial Ethernet protocols like EtherCAT, PROFINET, and EtherNet/IP require strict conformance testing. Certification labs validate behavior under edge cases such as fault conditions, recovery timing, and protocol‑specific safety functions. These processes can take weeks or months and represent a significant cost for each product variant. By contrast, the netX90’s pre‑certified stacks shift much of this burden to Hilscher. OEMs still need to perform product-level interoperability and regression tests, but they are no longer responsible for certifying the protocol stack itself. This distinction shortens test cycles and lowers cost while reducing risk.
How DE Design Works Helps
As a certified Hilscher integration partner, DE Design Works is one of the firms that Hilscher actively steers customers to when designing products with the netX90. Hilscher focuses its resources on developing SoCs and protocol stacks rather than providing direct engineering services, so they rely on partners like DE to help OEMs integrate the technology successfully.
Many engineering teams recognize the need for migration but underestimate the technical hurdles. DE Design Works brings:
- Experience with netX90 internals: We’ve ported real-world applications to netX90, including adapting ADIs and redesigning memory maps.
- Protocol expertise: Our engineers are familiar with EtherCAT, PROFINET, and EtherNet/IP certification processes.
- Hardware/software co-design: We evaluate both board layout and firmware requirements to ensure smooth integration.
- Structured methodology: Assessment → Architecture Mapping → Prototype → Validation, reducing risk at each step.
A client came to DE after receiving an EOL notice on a critical Anybus module. We migrated their STM32-based design to a netX90 platform, re-architecting firmware and adapting ADIs. The project began with a detailed assessment of memory usage and protocol requirements, followed by prototype bring-up and iterative validation. The outcome was a smaller, more efficient board that supported multiple protocols via licensing rather than hardware redesign. The client reported a 15% reduction in BOM cost, months saved in compliance testing, and gained increased confidence in their roadmap to support future protocols.
Conclusion
The move from STM32/Anybus to next-gen SoCs like the netX90 is more than a cost-saving measure. It is a technical modernization that improves performance, simplifies design, and positions products for long-term success. The advantages are clear: tighter integration, faster communications, reduced power and size, easier protocol switching, and stronger long-term maintainability. Together these factors make a compelling case for moving beyond legacy architectures.
Looking ahead, integrated SoCs like the netX90 align with Industry 4.0 and Industrial IoT trends. As security requirements tighten and demand grows for edge intelligence, designs that start with modern, integrated platforms will be better positioned to add features like remote updates, predictive maintenance, and cloud connectivity. Security is becoming especially important, with new regulations in industrial sectors requiring encrypted communications and robust update mechanisms. Companies that migrate sooner will have a strategic advantage.
For engineering leaders, the question is not if to migrate, but when and how to make the transition efficiently and with minimal risk. That’s where expertise makes the difference.
If you’re considering a migration project, DE Design Works is ready to help. Our team has hands-on experience with STM32-to-netX90 transitions, industrial protocol certification, and hardware/software co-design. Let’s start a conversation about how we can accelerate your next-generation platform and give your products a competitive edge.