AMD Ryzen AI Embedded X100 brings Zen 5 power to autonomous robots and physical AI applications

AMD's new X100 chip lineup puts Strix Halo into robots - APUs for physical AI bring Zen 5 CPU, RDNA 3.5 GPU cores to compete with Intel's Panther Lake

AMD is bringing its Strix Halo APUs into the world of physical AI. The new X100 series of processors come with similar specs as the various Ryzen AI Max models floating around in client devices, but they're tailored for 24/7 operation, with a 10-year lifecycle in embedded applications like robotics. There are three SKUs that align with the three original Strix Halo models (not the updated versions with 40 CUs). The top-end X199 comes with 16 Zen 5 cores and 40 RDNA 3.5 CUs. The X188 steps down to 12 cores and 32 CUs, while the X168 comes with eight cores and the same 32 CUs. AMD hasn't shared detailed specifications for each model, but the company says the range goes up to a 5.1 GHz boost clock and 128 GB of unified memory. They also include an XDNA 2 NPU with up to 50 TOPS, a configurable TDP between 45W and 120W, and operating temperature between -40 degrees Celsius up to 105 degrees. AMD's range bites back at Intel, which launched a range of Panther Lake SoCs for physical AI at the beginning of the year. Both make an argument for SoCs in robotics, reducing latency when the CPU, AI accelerator, and memory are fragmented across separate chips. The X100 range is just physically larger than Panther Lake, packing much more silicon on the SoC for more powerful deployments. The company shared a range of benchmarks comparing the flagship X199 against Intel's Core Ultra X7 358H, a 16-core chip with Intel's Arc B390 iGPU that has 12 Xe3 cores. AMD claims a lead of 1.2X and 1.3X, respectively, in GeekBench 6.1 and PassMark, as well as 1.5X in an unofficial SPECrate 2017 run looking at integer workloads. In graphics, AMD unsurprisingly takes the edge with 1.4X faster Vulkan and 1.7X faster OpenGL performance (both measured with GFXBench 5 on Ubuntu), as well as a 1.6X lead in Unigine Heaven Extreme. On the physical AI front, AMD claims a 1.4X improvement in Time to First Token (TTFT) and 3.5X faster tokens per second in Llama-bench, with a Vulkan backend running at a 45W TDP. These results need a massive dash of salt, however. AMD tested the Ryzen AI Max 395+ "configured to reflect Ryzen AI Embedded X199 specifications." It tested on the Maple reference board with a 5.1 GHz CPU clock, 2.9 GHz GPU clock, and sustained 45W TDP. The X7 358H, meanwhile, was tested in an MSI Prestige 16 Flip AI+ with an enforced TDP limit of 30W. AMD then "projected" 45W performance on the Intel chip "using scaling factors derived from public benchmark data." It's not exactly an apples-to-apples comparison, in other words. There's some sort of proxy stand-in or extrapolation of data across all of the benchmarks here, so keep that in mind as we work through the rest of AMD's X100 announcements. AMD X100 Kria SOM and robotics developer platform utside of the chips themselves, AMD is offering X100 models as part of a Kria System on Module (SOM) or an integrated robotics developer platform. The Kria X100 board measures 120mm x 120mm and conforms to the standardized COM-HPC form factor. If you're a developer that wants to develop for the board, AMD is offering its Kria AI robotics developer platform. It's a fully-integrated box, leveraging the X100 Kria SOM alongside AMD's Spartan UltraScale+ FPGA baseboard. AMD says it's a "turnkey" solution for robotics development, including specialized connectivity for cameras and industrial networking, along with robotic sensors. The platform is available in early access now, and AMD says it'll be in full production in Q4 of this year. . AMD shared some benchmarks for the X100 Kria, as well, comparing it to Nvidia's Thor T5000. These benchmarks weren't run internally at AMD. They were commissioned by AMD and ran by Open Navigation and Mimix. Critically, the benchmarks didn't test an X100 Kria board, or at least, not exactly in the form it will take once it's inside a robot or AMD's developer box. Instead, AMD is comparing Nvidia's Jetson AGX Thor developer kit to a GMKtech EVO-X2 AI mini PC with a Ryzen AI Max+ 395 "configured to reflect Ryzen AI embedded x199 specifications." Naturally, the thermal and power environment of these chips will heavily influence performance. AMD is continuing its attempt to siphon developers away from Nvidia's CUDA platform for development, as well. It's HIPIFY tool converts CUDA code to AMD's HIP C++ portable code, and the company claims it can now handle 70-80% of the "effort" of porting on its own. AMD tested on a Ryzen AI Max+ 395, once again configured to match the X199, and it ported 15 CUDA applications, comprising 1,199 lines of code, to arrive at that 70% to 80% range. X100 Kria lives at the "brain" of the robotics platform, but AMD envisions an end-to-end solution for humanoid-style robots with its Spartan UltraScale+. Zynq UltraScale+, and Versal AI Edge Gen 2 FPGAs and SoCs Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

SAPPHIRE Unveils EDGE+ Apex Robotics Platform Powered by AMD Ryzen AI Embedded X100

SAPPHIRE Technology has introduced the EDGE+ Apex SOM/Carrier Robotics Platform, a production-ready embedded hardware solution designed for autonomous robotics and physical AI applications. Demonstrated during AMD Advancing AI 2026 in San Francisco, the platform combines AMD's newly announced Ryzen AI Embedded X100 Series processors with a dedicated robotics carrier board to provide a complete edge computing solution for industrial automation, machine vision, and autonomous systems. The platform is built around a COM-HPC Client Type-B System-on-Module powered by AMD Ryzen AI Embedded X100 Series processors and supports up to 128 GB of LPDDR5x memory. The processor combines Zen 5 CPU cores, integrated RDNA graphics, and an XDNA 2 neural processing unit, allowing AI inference, graphics processing, and real-time control workloads to execute on a unified memory architecture. According to AMD, the architecture is intended to deliver deterministic performance for robotics while maintaining compatibility with standard x86 software environments rather than relying on proprietary AI hardware. SAPPHIRE complements the compute module with a robotics-oriented carrier board built around an AMD UltraScale+ FPGA. The carrier provides a broad range of industrial interfaces including GMSL camera inputs with integrated power delivery, CAN-FD, EtherCAT and TSN networking, multi-gigabit Ethernet through QSFP, USB4 Type-C, PCI Express Gen5 expansion via OCuLink, and an integrated inertial measurement unit (IMU). Together, these interfaces allow developers to connect cameras, sensors, actuators, storage, and industrial networking equipment without requiring additional expansion hardware. The platform also incorporates several features aimed at industrial and safety-critical deployments. These include a Functional Safety (FuSa)-capable power subsystem, Trusted Platform Module (TPM) 2.0 support for secure boot and encryption, and long-term embedded lifecycle support. These characteristics make the platform suitable for continuous operation in manufacturing, logistics, healthcare, aerospace, defense, and autonomous mobile robotics. Software support focuses on an open development ecosystem. Developers can utilize established AI frameworks while taking advantage of AMD ROCm alongside Ryzen AI software to distribute workloads across the CPU, GPU, and NPU. This approach enables hardware acceleration for perception, object recognition, planning, and autonomous decision-making without requiring developers to migrate to proprietary software stacks. SAPPHIRE positions EDGE+ Apex as a development platform intended to shorten the path from prototype to commercial deployment. Rather than requiring customers to integrate individual compute modules, FPGA hardware, sensor interfaces, and industrial connectivity themselves, the company provides a pre-engineered hardware foundation optimized specifically for robotics workloads. Specifications FeatureSpecification ProcessorAMD Ryzen AI Embedded X100 Series Module TypeCOM-HPC Client Type-B SOM MemoryUp to 128 GB LPDDR5x AI AccelerationZen 5 CPU, RDNA GPU, XDNA 2 NPU Carrier FPGAAMD UltraScale+ Camera InterfacesGMSL with integrated power Industrial NetworkingCAN-FD, EtherCAT, TSN EthernetMulti-Gigabit QSFP ExpansionOCuLink PCIe Gen5 USBUSB4 Type-C Integrated SensorsOnboard IMU SecurityTPM 2.0 Safety FeaturesFunctional Safety (FuSa)-capable power design Target MarketsRobotics, machine vision, industrial automation, physical AI, healthcare, aerospace, defense

AMD's Ryzen AI Embedded X100 Series processors are built for autonomous robots and physical AI

At its annual Advancing AI 2026 event, AMD has not only launched its next-gen Zen 6-powered AMD EPYC 'Venice' CPUs and AMD Instinct MI455X GPUs for large-scale AI, but it has also announced the new AMD Ryzen AI Embedded X100 Series processors for physical AI and robotics. These new processors combine a 16-core Zen 5 CPU with a discrete-class integrated RDNA 3.5 iGPU with 40 Compute Units, similar to AMD's Strix Halo processor, the Ryzen AI Max+ 395. For physical AI and robotics, the 16-core CPU covers things like planning, orchestration, and agentic AI control, while the GPU handles perception, reasoning, and vision. Like other Ryzen AI chips, there's an onboard XDNA 2 NPU, which handles continuous vision and audio. The Ryzen AI Embedded X100 Series processors power AMD's new Kria AI system-on modules (SOMs), so developers and designers can easily prototype, build, and deploy autonomous robots and physical AI systems. In fact, the AMD Kria AI Robotics Developer Platform is the first "open, turnkey integrated platform for autonomous robotics" thanks to its combination of CPU, GPU, NPU, and FPGA. AMD notes that when stacked up against the latest Intel Core Ultra Series 3 processors, you've got up to 2.1x higher multithread CPU performance, 1.7x graphics performance, 3.5x higher token generation throughput, and 1.4x faster time-to-first-token (TTFT). That last one is important as it helps accelerate physical AI workloads. And it's supported by an open software stack for Linux with AMD ROCm and support for AI frameworks like PyTorch, ONNX and TensorFlow. And as processors designed for autonomous robots, they're built to survive and operate in harsh conditions including sub-zero temperatures of -40 to 105 degrees Celsius. Not only that, but AMD says that they can run 24/7 for 10 years in harsh environments. AMD has begun sampling its new Ryzen AI Embedded X100 Series Processors to customers, with a full release expected in Q4 2026.

AMD Ryzen AI Embedded X100 Combines 16-Core Zen 5 CPU, GPU and NPU

AMD has introduced the Ryzen AI Embedded X100 Series for robotics, industrial automation, medical equipment and other physical AI systems. The processors combine up to 16 Zen 5 CPU cores, an integrated Radeon GPU and an NPU within a single embedded SoC. Physical AI refers to systems that receive information from their environment, process it and generate a physical response. Examples include industrial robots, autonomous machines, medical-imaging equipment and unmanned platforms. These applications may need to perform AI inference while simultaneously handling sensor input, graphics, signal processing and real-time control software. The CPU, GPU and NPU operate through a unified memory architecture. Sharing one memory pool can reduce the number of transfers required as information moves between processing stages. A machine-vision application, for example, could capture and prepare an image, perform neural-network inference and send the result to its control software without copying the complete data set between separate CPU and accelerator memory spaces. AMD has supplied several internal performance comparisons. The company expects up to 2.1 times higher multithreaded CoreMark performance and 1.7 times higher OpenGL graphics performance than an Intel Core Ultra X7 358H. For local language-model inference, AMD reports 3.5 times higher token-generation throughput and 1.4 times faster time to first token. The tests covered selected Gemma, Llama and Qwen models requiring less than 24 GB of memory. These figures were not produced with final X100 hardware. AMD used a Ryzen AI Max+ 395 configured to approximate the Ryzen AI Embedded X199. Some of the Intel results were also projected to a 45-watt operating level from measurements performed at 30 watts. The results should therefore be considered platform estimates rather than independent comparisons of production processors. AMD separately claims up to three times the peak FP32 performance of Nvidia's Jetson T5000. In a medical-ultrasound beamforming workload, an AMD proxy platform reportedly completed processing an average of 1.7 times faster than a system containing an Nvidia RTX 4000 SFF Ada GPU. The AMD platform in that comparison used a Ryzen AI Max+ PRO 395, Radeon 8060S graphics and 128 GB of unified LPDDR5X-8000 memory. The competing system combined a Ryzen 7 9800X3D with an RTX 4000 SFF Ada carrying 20 GB of GDDR6. Differences in memory capacity and architecture should be considered when interpreting the result. Software support includes Linux, ROCm and Xen Hypervisor, together with PyTorch, ONNX and TensorFlow. AMD will also provide tools for migrating some CUDA applications to ROCm, although software using CUDA-specific libraries may still require manual porting and optimization. The X100 Series is rated for operation between -40°C and 105°C. AMD also specifies continuous 24/7 operation and availability for up to ten years. Customer sampling began in June 2026, with production scheduled for the fourth quarter. System-on-module partners include Arbor, Congatec, iBase, IEI, Sapphire and Seavo. Specification AMD Ryzen AI Embedded X100 Series Processor type Embedded system-on-chip CPU architecture AMD Zen 5 Maximum CPU configuration Up to 16 cores Integrated graphics AMD Radeon GPU AI accelerator Integrated NPU Memory architecture Unified memory shared across CPU, GPU and NPU Operating-temperature range -40°C to 105°C Operating profile Continuous 24/7 operation Planned availability Up to ten years Operating-system support Linux Compute software AMD ROCm Virtualization Xen Hypervisor AI frameworks PyTorch, ONNX and TensorFlow Target applications Robotics, industrial automation, healthcare, aerospace, defense and unmanned systems Customer sampling Started June 2026 Production availability Expected in Q4 2026 Module partners Arbor, Congatec, iBase, IEI, Sapphire and Seavo Source: AMD