New SOSA-Aligned FPGA PIC Module for Tactical Sensor Processing Applications

Curtiss-Wright is unveiling the VPX3-536 Adaptable Processor, a new SOSA-aligned 3U VPX FPGA plug-in-card (PIC) module designed for demanding tactical sensor processing applications such as radar and electronic warfare By William Mackenzie / 14 Feb 2025

Single Board Computer (SBC)

Discover cutting-edge solutions from 5 leading global suppliers
SUPPLIER SPOTLIGHT
Follow DA

Curtiss-Wright’s Defense Solutions Division is releasing a new high-performance SOSA-aligned 3U VPX FPGA plug-in-card (PIC) module, the VPX3-536 Adaptable Processor

The VPX3-536, the newest addition to Curtiss-Wright’s Fabric100™ ecosystem of 100 Gigabit Ethernet (GbE)/PCIe Gen 4 system modules, combines an AI-enabled accelerated compute architecture, dual 400G high-speed crypto engines, and up to 28 high-speed backplane fiber optic links. 

New SOSA-Aligned FPGA PIC Module for Tactical Sensor Processing Applications

The new module is designed for use in demanding sensor processing applications, such as latency-sensitive radar, electronic warfare, and SIGINT digital signal processing (DSP).

The VPX3-536 is a size, weight, and power (SWaP) optimized which can securely and simultaneously ingest, process, and egress up to 784 Gbps of optical sensor or Ethernet data over the system backplane (100 GbE/Serial/Aurora interfaces) using high-density, high-performance VITA 66.5 interfaces (28 lanes/full-duplex/28 Gbps).

The flexible, user-programmable board features AMD’s highest performance heterogeneous Adaptive Compute Acceleration Platform (ACAP) architecture. The AMD Versal™ Premium VP2502 Adaptive System-on-Chip (SoC) integrates a large FPGA array (3.7M logic cells), two dual-core Arm® processors, and powerful DSP engines (7,392 slice) to eliminate the need for a daughtercard or a discrete processor card to manage it. 

This Versal device provides 472 Adaptive Intelligence Engines (157 TOPS) to support AI, Machine Learning (ML), and 5G signal processing. Overall application performance can reportedly be enhanced by splitting signal processing workloads between the board’s DSP and AI engines. The module’s built-in crypto engine supports line-rate MACSec, IPSec and bulk encryption.

The module’s AMD (formerly Xilinx) Versal Premium ASoC device enables users to leverage and protect their use of Xilinx architecture solutions. With its low latency Network on Chip (NoC) and powerful multi-core Arm processor architecture, the VPX3-536 can serve as a standalone, single-slot ‘system on a board’ running the PetaLinux operating system.

It can also be used as a co-processing resource in multi-board embedded VPX systems with other system elements such as Curtiss-Wright’s CHAMP-XD3 Intel® processor and Fabric100 VPX3-1262 Intel processor and VPX3-6816 Network Switch modules.

Brian Perry, Senior Vice President and General Manager, Curtiss-Wright Defense Solutions Division, commented, “Curtiss-Wright is proud to introduce a truly game-changing heterogeneous architecture FPGA engine that unleashes unmatched compute power and throughput for deployed sensor processing applications. No other SOSA aligned 3U VPX card on the market offers the breakthrough combination of FPGA, CPU, DSP processing, AI acceleration and fiber optic backplane throughput delivered by our new VPX3-536 module.”

Posted by William Mackenzie Edited by William Mackenzie, Junior Editor and Copywriter practised in digital content creation and writing about defense and marine technology. Connect & Contact

Latest Articles

C5ISR & the Evolution of Situational Awareness in Defense Operations

Elistair explores how C5ISR integrates command, intelligence, and combat systems to enhance situational awareness and decision-making in modern military operations

Jul 16, 2025
RF Drone Detection Solution Selected for Integration into C-UAS Capability

Metis will supply a NATO consumer with its Skyperion Radio Frequency (RF) drone detection components for defending airspace, infrastructure and installations against rogue drones

Jul 16, 2025
Static Calibration Techniques for Silicon Designs Accelerometers

Learn how to perform quick, accurate static calibration of Silicon Designs MEMS DC accelerometers using only a voltmeter and gravity, no specialized equipment required

Jul 16, 2025
Disasters Expo USA & Resilient City Expo Co-Locating in Houston

Disasters Expo USA and Resilient City Expo will take place on November 5–6, 2025, bringing professionals to Houston to address crisis response and long-term resilience

Jul 15, 2025
General Digital’s Modular Approach to Rugged Display Design

General Digital outlines how its flexible design approach delivers rugged, customer-tailored military displays, rigorously tested in-house even when formal MIL-STD certification is not required

Jul 15, 2025
Precision Underwater Sensing for Critical Naval & Security Missions

Defense Advancement showcases Impact Subsea's underwater sensing technologies for naval and security missions

Jul 15, 2025

Featured Content

Q&A: How Revector Supports Defense with Mobile Intelligence

This Q&A provides insight into Revector’s technology, detailing its role in covert surveillance, threat detection, and integration with defense and security systems

Jul 14, 2025
New Oscilloquartz Optical Cesium Clocks Enhance Network Timing Precision

Adtran is expanding its Oscilloquartz portfolio with the new OSA 3200 SP and OSA 3250 ePRC optical cesium models, providing improved holdover and simplified maintenance for telecom, data center, utility, and defense networks

Jul 09, 2025
Teledyne FLIR OEM Whitepaper: Synthetic Data Generation for AI Object Detection

Read Teledyne FLIR OEM’s whitepaper on AIMMGen™, a defense-oriented synthetic data and AI model generation pipeline tailored to Automatic Target Recognition (ATR) and object classification tasks

Jul 03, 2025
Advancing Defense Capability Through Strategic Collaboration Defense Advancement works with major OEMs to foster collaboration and increase engagement with SMEs, to accelerate innovation and drive defense capabilities forward.