What you’ll learn:
- What does the X100 development platform look like?
- Why is this using the industry-standard COM-HPC form factor?
AMD’s Ryzen AI Embedded X100 processor (APU) targets applications that require AI support and real-time capabilities, such as robotics. I talked with Rob Bauer, Senior Manager in the Embedded Business Unit, about the new processor and the Kria AI Robotics platform (Fig. 1). Inside is a COM-HPC standard module.
The development system includes PC-style connections on one side, including USB and Ethernet connections, while the other end focuses on peripheral connectivity like camera interfaces.
The AMD Kria AI platform is built around the AMD X100, which is based on TSMC’s 4-nm technology (Fig. 2). The CPU complex includes 16 Zen 5 cores with AVX-512 support and simultaneous multithreading, while the integrated RDMA 3.5 GPU includes 40 compute units. The neural processing unit (NPU), based on AMD’s XDMA 2 architecture, is designed for low-power, always-on operation. The low-power NPU can deliver 50 TOPS of performance.
AMD’s software support is based on the open-source ROCm.AI (Fig. 3). This includes the robotics core SDK with ROS 2 support. The company has its own rocLibrary runtime and it supports the HIP (Heterogeneous-compute Interface for Portability) API. HIP is a thin interface layer for rocLibrary. The HIPIFY tool assists in the conversion of CUDA code to HIP code.
The AI and robotics software target any of the processing complexes from CPU to NPU. This allows the NPU to handle some AI models using minimal power, and the GPU can be utilized when there’s a need for more computational power.
The COM-HPC module is available separately, enabling prototypes to be quickly transferred to production. Supporting up to 128 GB of LPDDR5x memory, it’s designed for an industrial temperature range of −40 to +85°C. Schematics and support documents for the open-source AMD Kria AI platform’s carrier board are available to developers.
About the Author
William G. Wong
Senior Content Director - Electronic Design and Microwaves & RF
I am Editor of Electronic Design focusing on embedded, software, and systems. As Senior Content Director, I also manage Microwaves & RF and I work with a great team of editors to provide engineers, programmers, developers and technical managers with interesting and useful articles and videos on a regular basis. Check out our free newsletters to see the latest content.
You can send press releases for new products for possible coverage on the website. I am also interested in receiving contributed articles for publishing on our website. Use our template and send to me along with a signed release form.
Check out my blog, AltEmbedded on Electronic Design, as well as his latest articles on this site that are listed below.
You can visit my social media via these links:
- AltEmbedded on Electronic Design
- Bill Wong on Facebook
- @AltEmbedded on Twitter
- Bill Wong on LinkedIn
I earned a Bachelor of Electrical Engineering at the Georgia Institute of Technology and a Masters in Computer Science from Rutgers University. I still do a bit of programming using everything from C and C++ to Rust and Ada/SPARK. I do a bit of PHP programming for Drupal websites. I have posted a few Drupal modules.
I still get a hand on software and electronic hardware. Some of this can be found on our Kit Close-Up video series. You can also see me on many of our TechXchange Talk videos. I am interested in a range of projects from robotics to artificial intelligence.
Facts Only
AMD developed the Ryzen AI Embedded X100 processor (APU).
The X100 is manufactured using TSMC’s 4-nm technology.
The CPU complex contains 16 Zen 5 cores with AVX-512 support and simultaneous multithreading.
The integrated RDMA 3.5 GPU contains 40 compute units.
The XDMA 2 architecture-based NPU provides 50 TOPS of performance.
The hardware utilizes the COM-HPC standard form factor.
The system supports up to 128 GB of LPDDR5x memory.
The operating temperature range is -40 to +85°C.
Software support is provided via open-source ROCm.AI and the HIP API.
The HIPIFY tool converts CUDA code to HIP code.
The Kria AI Robotics platform includes a carrier board with open-source schematics.
Executive Summary
AMD has introduced the Ryzen AI Embedded X100, an APU specifically engineered for robotics and real-time AI applications. By integrating a 16-core Zen 5 CPU, a 40-compute-unit GPU, and a 50 TOPS NPU on a 4-nm process, the hardware allows for scalable power management; low-power tasks run on the NPU, while computationally intensive workloads shift to the GPU.
The platform emphasizes industrial viability and developer flexibility. Utilizing the COM-HPC standard form factor ensures that prototypes can transition quickly to production. To lower the barrier for software migration, AMD provides the HIPIFY tool to translate CUDA code into HIP, supported by the open-source ROCm.AI ecosystem and ROS 2. While the hardware specifications are precise, the actual real-world performance in diverse robotics environments remains to be demonstrated.
Full Take
The strongest version of this narrative is that AMD is strategically positioning itself to challenge the dominance of proprietary AI ecosystems in the robotics sector by offering a high-performance, open-standard alternative. By combining industrial-grade hardware (COM-HPC) with tools to migrate away from CUDA, they are attempting to reduce vendor lock-in.
However, this content functions as a vendor-driven showcase. The primary persuasive vector is the presentation of high-specification numbers (50 TOPS, 16 Zen 5 cores) as a proxy for capability, while the actual efficacy of the "open-source" software layer in complex, real-world robotics deployments is left unexamined. The narrative relies on the assumption that hardware specs and API translation tools are the primary bottlenecks in robotics, rather than the systemic challenges of sensor fusion or real-time OS stability.
Patterns detected: ARC-0063 Authority Game
The driving paradigm is the "commodity-AI" push: the belief that AI capability can be solved through silicon density and API compatibility. The implication is a shift in agency from software architects toward hardware providers who control the underlying "accelerator" logic.
Bridge Questions: How does the actual power efficiency of the XDMA 2 NPU compare to competing ARM-based edge AI solutions? Does the HIPIFY tool maintain performance parity, or does it introduce latency that could be critical in real-time robotics?
Counterstrike Scan: A coordinated campaign would use "open-source" rhetoric to mask a transition from one proprietary ecosystem (Nvidia) to another (AMD) while inflating TOPS numbers to create a sense of inevitable obsolescence. The content aligns partially with this by framing a product launch as a technical briefing, though it lacks the aggressive fear-mongering typical of a full influence operation.