embedded world 2025 put practical edge computing in the spotlight: Arm demonstrated local AI on a Raspberry Pi 5, Altera and GOWIN connected programmable logic with AI and industrial control, and Qualcomm and DigiKey emphasized development kits, reference designs and prototyping. The event ran in Nuremberg from 11–13 March 2025. These were announcements and demonstrations, not proof of comparative performance or broad product availability.
What was embedded world 2025?
The Nuremberg exhibition ran from 11–13 March 2025. The organiser described its 23rd edition as spanning seven exhibition halls, alongside the embedded world Conference and the electronic displays Conference. Ahead of the event, the organiser highlighted Altera CEO Sandra Rivera’s keynote, “Flexible AI at the Edge.”
In its 2025 post-event release, embedded world and NürnbergMesse reported almost 1,200 exhibitors from 46 countries and around 32,000 visitors from more than 80 countries. The conferences drew 1,897 participants from 47 countries. These are the organiser’s event figures.
What did the event show about edge AI?
Arm demonstrated a local AI workflow on Raspberry Pi 5
Arm’s booth demonstration paired an Arm-powered Raspberry Pi 5 with YOLO11 for object detection and the TinyStories small language model (SLM) to generate a story. Arm said the processing ran on the device. That makes the demonstration a concrete example of an edge-AI workflow: a device analyzes input and produces an output locally, rather than relying on a cloud service for those tasks.
#1 Best Overall
- Designed for students and beginners looking to understand Digital Logic, fundamentals of FPGAs
- Features the Xilinx Artix 7 FPGA compatible with Vivado Design Suite WebPACK Edition (free download available from Xilinx)
- On board user interfaces include 16 user switches, 16 LEDs, 5 user pushbuttons, and a
- Expansion opportunities with four Pmod ports including 3 standard 12-pin Pmod ports and 1 dual
- Does NOT ship with micro USB cable
The demo illustrates an approach, not a published performance benchmark. Arm’s recap does not establish model latency, accuracy, power consumption, memory use, or whether the setup was a complete retail bundle. It also does not mean every Raspberry Pi 5 project can run the same models without configuration or trade-offs.
Qualcomm emphasized development and sector-specific use cases
Qualcomm described developer kits and demonstrations for intelligent computing in automotive, healthcare and IoT. Its event announcement placed the demonstrations at booth 5-161. The emphasis was on development kits, reference designs and applications—ways for engineers to explore deployment—rather than only announcing silicon. The announcement does not provide comparable benchmark figures, pricing or shipping details for the kits.
Rank #2
- Arty A7 comes in two FPGA variants: Arty A7-35T features Xilinx XC7A35TICSG324-1L. Arty A7-100T features the larger Xilinx XC7A100TCSG324-1.
- Internal clock speeds exceeding 450MHz, On-chip analog-to-digital converter (XADC), Programmable over JTAG and Quad-SPI Flash
- 256MB DDR3L with a 16-bit bus @ 667MHz, 16MB Quad-SPI Flash, USB-JTAG Programming circuitry, Powered from USB or any 7V-15V source
- 10/100 Mbps Ethernet, USB-UART Bridge
- 4 Switches, 4 Buttons, 1 Reset Button, 4 LEDs, 4 RGB LEDs, 4 Pmod connectors, shield connector
How were FPGAs positioned for AI and industrial systems?
Altera announced an FPGA software-and-hardware stack
Altera announced its latest Agilex FPGAs, Quartus Prime Pro software and FPGA AI Suite for customized embedded systems. It identified robotics, factory automation and medical equipment as target applications, framing the offering around the power, performance and size constraints of embedded designs. The announcement describes a toolchain and product family, but does not provide enough detail to compare specific device performance, availability or pricing with other platforms.
GOWIN demonstrated motor control and video bridging
GOWIN showed FPGA-based motor-control and video-bridging designs. Its GW5AS motor-control demonstration used a GW5AS-25K solution that combines an Arm Cortex-M4 processor rated up to 288 MHz with an Arora-V FPGA containing 25K LUTs. This is a heterogeneous design: a processor and programmable logic share a system, enabling designers to divide control and hardware-specific tasks between them. The cited demonstration does not establish motor-control performance under a stated test protocol.
Rank #3
- [FPGA Chip] GW2AR-18 QN88 FPGA Chip containing 20736 LUT4 logic cells and 15552 Filp-Flops.There are 2 PLL in this FPGA chip, and many DSP units supporting 18 bit x 18 bit multiplication
- [Onboard Debugger ] Sipeed Tang Nano 20K Development Board support JTAG for FPGA, USB to UART for FPGA,USB to SPI for FPGA communication, Control MS5351 generate frequency
- [USB2.0 HS interface] The 27MHz crystal generates the clock for HDMI display, onboard MS5351 clock generating chip also provides mutiple clocks.Support Serial communication, high-speed SPI reception.
- [Application scenarios] Tang Nano 20K Open source Development Board supports game console emulators, drives RGB screens, multiple display outputs, 20K LUT4, RISC-V soft-core experiments.
- [Wiki] "dl.sipeed.com/shareURL/TANG/Nano_20K/1_Datasheet";Any after-Sales Privems, Please Contact us by click "Waypondev" store and ask a question or leave the message in our forum by "forum.youyeetoo .com/".
What could engineers prototype or explore?
DigiKey connected components with remote and sensor projects
DigiKey said its booth would feature LabsLand Prism4, a modular ecosystem for interactive, real-time remote hardware systems, and an optical heart-rate sensing demonstration using NXP’s MCX N94 microcontroller. These examples represented the prototyping and component side of the exhibition: remote access to hardware systems and a microcontroller-based sensing application. The announcement does not specify a retail package, price or performance results for either demonstration.
Compare the demonstrations by what they make possible
| Example | Architecture | What was shown or announced | Target or use | Availability or performance details in the cited announcement |
|---|---|---|---|---|
| Arm and Raspberry Pi 5 | Arm-powered single-board computer | YOLO11 object detection followed by TinyStories SLM story generation; Arm said processing ran on-device | Local AI experimentation | Not stated (Arm recap) |
| Altera Agilex, Quartus Prime Pro and FPGA AI Suite | FPGA hardware with development software and AI tools | Announced as a stack for customized embedded systems | Robotics, factory automation and medical equipment | Specific benchmark, pricing and shipping details not stated (Altera announcement) |
| GOWIN GW5AS-25K motor-control demo | Heterogeneous: Cortex-M4 processor plus Arora-V FPGA | Motor-control design; processor rated up to 288 MHz and FPGA with 25K LUTs | Industrial motor control; GOWIN also showed video bridging | Demo performance under a stated test protocol not stated (GOWIN announcement) |
| Qualcomm developer kits and demos | Not stated in the event announcement | Developer kits and demonstrations for intelligent computing | Automotive, healthcare and IoT | Specific kit specifications, pricing and shipping details not stated (Qualcomm announcement) |
| DigiKey booth demonstrations | LabsLand Prism4 remote hardware ecosystem; NXP MCX N94 microcontroller for optical heart-rate sensing | Interactive real-time remote hardware systems and a sensing demo | Prototyping and embedded sensing | Specific package, pricing and performance details not stated (DigiKey announcement) |
The clearest distinction is architectural. Raspberry Pi 5 offered a visible example of running an AI workflow on a general-purpose board; FPGA platforms target configurable hardware, including systems that combine a processor and programmable logic; and DigiKey’s examples focused on remote hardware access and microcontroller-based sensing. The event materials do not provide a common test or enough power, latency, software-maturity or cost data to rank these approaches. A choice for a real design depends on its workload, constraints and available development support.
Quick Recap
Best Value
- Digilent Basys 3 Artix-7 FPGA Trainer Board: Recommended for Introductory Users
Rank #4
- The best way to get started with FPGAs: Using a simple board with projects that build on eachother, now anyone can get started with FPGA development!
- Fun peripherals available: With 4 LEDs, 4 push-buttons, 7-segment display, USB connector, a VGA connector, and a PMOD (for expansion) you can have dozens of fun projects available to you out of the box!
- Works with Verilog and VHDL: No matter which programming language you want to get started with, the Go Board will work for you!
- No extra device required: Simply plug the Go Board into a USB port and go! Getting started with FPGAs has never been easier.
- Works with all operating systems: Windows, Mac, Linux
What should readers take away from the announcements?
- Edge AI was demonstrated as a complete workflow. Arm’s Raspberry Pi 5 example joined object detection and text generation on one device, moving beyond a generic claim that AI can run at the edge.
- Programmable logic was presented as part of embedded AI and control. Altera announced an FPGA-oriented stack for customized systems, while GOWIN’s motor-control demo showed a processor-plus-FPGA configuration.
- Development pathways mattered alongside chips. Qualcomm highlighted kits and reference-style demonstrations, and DigiKey showed remote hardware and sensor prototyping concepts.
- Demonstrations are not a substitute for product validation. Vendor announcements establish what was presented or announced at the event; they do not, on their own, establish independent benchmark results, shipping schedules, prices or long-term adoption.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
