AMD Ryzen AI Embedded X100 targets robotics with up to 128GB unified memory
AMD has launched the Ryzen AI Embedded X100 series at Advancing AI 2026. The processors use the same Strix Halo silicon found in the Ryzen AI Max series, but target robotics, industrial automation, healthcare and other embedded systems.
AMD previewed the X100 family at CES 2026 without confirming the full product range. The final lineup includes six processors, led by the Ryzen AI Embedded X199 with 16 Zen 5 cores and 40 RDNA 3.5 compute units. The X188 has 12 cores and 32 CUs, while the X168 reduces the CPU configuration to eight cores but retains the 32-CU GPU.
Strix Halo moves into embedded systems
Each processor includes an XDNA 2 NPU rated at up to 50 TOPS. The chips have a default TDP of 55W and support a configurable range from 45W to 120W. Current COM-HPC modules support up to 128GB of LPDDR5X-8533 memory shared between the CPU, GPU and NPU.
© AMD 
© AMD The models carrying the “i” suffix are qualified for industrial temperature ranges. AMD rates the processors for continuous operation and product availability of up to ten years. The commercial models have a junction temperature range starting at 0°C, while the industrial variants are rated down to -40°C.
Production availability planned for Q4 2026
The X100 platform supports Linux, AMD ROCm, the Xen Hypervisor and AI frameworks including PyTorch, ONNX and TensorFlow. Customer sampling began in June 2026. Production availability is planned for the fourth quarter, with systems and modules expected from Arbor, congatec, iBase, IEI, Sapphire and Seavo.
© AMD AMD claims up to 2.1 times higher multithreaded CoreMark performance and 1.7 times higher OpenGL graphics performance than Intel Core Ultra Series 3. The company also reports 3.5 times higher token generation throughput. These results were not collected from final X100 hardware. AMD used a Ryzen AI Max+ 395 configured to match the X199 frequency and 45W sustained power settings.
Source: AMD
