Application Highlight: The Andromeda XRU50 SOM from Enclustra Incorporates AMD RFSoC

August 12, 2024

Sponsored Story

FPGA technology has simplified and streamlined embedded systems in many industries, allowing OEMs to shrink sizes, improve performance, and drive down the cost of electronics applications.

However, FPGA can require complex system design to incorporate into existing systems and connectivity spectra, especially in radiofrequency applications. This complexity of integration can offset or even eliminate many of the advantages that come from using an FPGA in other embedded systems.

One solution to that complexity is the RFSoC module. By incorporating this kind of module, built specifically for radio frequency management, the user gets access to what is basically a single SoC that incorporates both tools.

RFSoC devices can integrate an FPGA along with other components like RF front-end technologies, analog-to-digital converters (ADCs), and digital-to-analog converters (DACs). By integrating all these tools into one embedded device, manufacturers can eliminate the need for many components, including external data converters. All of which reduces part counts, eases software integration, and lowers both costs and power consumption. A win all the way around.

Application Use Case

Why is this even a concern? With all the advantages of FPGA, even if it’s a bit more complex to use in RF applications, is such an architecture as the RFSoC even needed?

Well, yes, depending as always on the application. For instance, consider spectrum monitoring. A major challenge in spectrum monitoring is maintaining fidelity as the band gets wider and wider. As users reach for more, and wider, frequency bands for their wireless signals, monitoring tools need to monitor ever-growing RF bandwidths. This requires very high sample rates, which means that an RFSoC with fast ADCs and enough FPGA fabric to handle massive parallel computations can solve this modern problem, even as it grows.

Similar problems arise and can be addressed with RFSoC in such applications as radar beamforming, telecommunications geofencing, advanced phased-array radar solutions, test and measurement, and in the future, even controlling quantum computers with their expected massive need for bandwidth.

Enclustra Andromeda XRU50 RFSoC

Enclustra announced at embedded world 2024 its Andromeda XRU50 RFSoC module that is designed specifically to bring flexibility and simpler integration to FPGA-based radiofrequency applications. The Andromeda XRU50 integrates AMD’s Zynq UltraScale+ RFSoC to support RF applications in 5G, phased-array radar solutions, and Quantum Computing.

It’s designed with eight digital-to-analog and analog-to-digital radiofrequency converters to enable direct synthesis of RF signals with no additional analog circuitry, the company said. The end goal for Enclustra seems to be to offer a complete, single-chip, software-defined radio platform.

“Demand for enhanced RF signal processing capabilities is soaring, particularly for size-constrained, decentralized solutions,” said Gaël Paul, VP of Innovation at Enclustra. “Our response to this trend is the Andromeda XRU50 RFSoC, which effectively streamlines the development of the next generation of sophisticated FPGA-based RF systems.”

The Andromeda XRU50 reportedly is built with programmable logic and highspeed data converters that make it possible to directly interface with the built-in FPGA’s programmable logic. Directly generating microwave signals with highspeed data converters allows the transmitter to achieve a signal bandwidth of 4.75 GHz, Enclustra says.

The XRU50 finds its real power in how it simplifies the integration of AMD’s Zynq UltraScale+ RFSoC into end products. The AMD platform combines FPGA fabric with 930,000 system logic cells, four Cortex-A53 1.3 GHz processors, two Cortex-R5F 533 MHz processors, and eight digital-to-analog and analog-to-digital radiofrequency converters, while the 80-by-64-millimeter XRU50 provides product developers easy access to the RFSoC’s more than 200 programmable logic I/O options, 20 multi-gigabit transceivers, and 32 GB of on-module memory. What’s more, the Gigabit ethernet and USB 3.0 support and PCIe and DDR4 memory interfaces make connectivity reliable and flexible.

The company says the XRU50 architecture is purpose-built for 5G wireless communication networks and facilitates the deployment of massive MIMO antennas and beamforming arrays. Radar systems need this kind of dynamic and adaptive beamforming to enhance performance enough for modern target tracking and identification.

Looking ahead at the realm of quantum computing, RFSoC offers high-speed, low-latency processing for precise microwave pulses that are essential for controlling the binary state of qubits in quantum computers.

The expansion of bandwidth and the demand for it seems inevitable, and embedded systems must begin to plan not just to handle that demand, but proactively prepare to capitalize upon it. An RFSoC is right now the best way to begin future-proofing embedded solutions, and preparing for opportunities and challenges of tomorrow, like the coming of Quantum Computing.

Even if that kind of forward-thinking isn’t in your roadmap, however, it’s worth remembering that smaller form factors, shorter time to market, faster processing, and lower cost will always be critical to customer retention and attraction, and incorporating FPGA most comprehensively will ease the challenge in making those offerings to the marketplace. So it’s still worth it to consider an RFSoC that comes with the FPGA integration built in.   

Additional Resources:

• Product Page: https://www.enclustra.com/en/products/system-on-chip-modules/andromeda-xru50/
• Order Form: https://www.enclustra.com/en/order/